Docstoc

Evolution of Photogrammetry Cross Currency Pairs

Document Sample
Evolution of Photogrammetry Cross Currency Pairs Powered By Docstoc
					COVER PAGE
                        ORGANISING COMMITTEE
1. Maj Gen (Dr.) B Nagarajan, Addl.Surveyor General,IIS&M   Chairman
                     & Chairman, Institution of Surveyors

2. Sh. Ashok Prim, DSG, Head Faculty TS&LIS, IIS&M          Executive Member

3. Sh. J K Rath, DSG, Head Faculty CDM&GIS, IISM            Executive Member

4. Sh. P V Srinivas, DSG, Head Faculty Geodesy, IISM        Executive Member

5. Sh.A Purnachandra, SS, Head Faculty P & RS, IISM         Executive Member

6. Sh A K Das, SS, Faculty of Geodesy, IIS&M                Member

7. Sh B K Mahapatra, SS, Faculty CDM&GIS, IISM              Member

8. Sh B Mahanta, SS, Faculty of TS&LIS, IIS&M               Member

9. Sh F M Panigrahi, SS, Faculty CDM&GIS, IISM              Member

10. Sh G C Nayak, SS Faculty of P & RS                      Organizing Secretary




                                          2
                                         PREFACE

       The Survey of India (SOI) , the National Mapping Agency, has been rendering its
service to the nation since centuries in development and security through providing the
geospatial data. The SOI has also been a pioneer in adopting the best available
technologies in the field of Surveying and Mapping maintaining its brand value Accuracy
and Standard. Indian Institute of Surveying & Mapping, the capacity building arm of SOI,
has been in the forefront in equipping the department, other national organizations and
Mapping agencies of neighboring countries to adopt the best possible technology at that
point of time.
       The century old technology of surface data acquisition from photographs is being
exploited by Surveying & Mapping fraternity to its maximum advantages. The technology
has gone through the various phases of development with the advancement of electronic
computation and data acquisition technologies from air and space. The Digital
Photogrammetry technology has progressed very fast in the last decade. Due to various
factors SOI could not keep pace with that to exploit it to its full potential while industries
have adopted it in large scale in their production line.
       Effective implementation of Digital Photogrammetry in the department is essential
to meet the national responsibility to provide updated geospatial data in required resolution
in a time bound manner for various applications. The Workshop organized jointly by
Indian Institute of Surveying & Mapping and Institution of Surveyors (Hyderabad Branch)
with the participation of professional from the department, other organisations and
industries will provide an opportunity to share the knowledge and experiences for effective
implementation.




                                               3
                                    CONTENTS


1. Back ground
2. Objectives of the workshop
3. Formation of working groups
3.1 Working Group I: Choice of Input Data
3.2 Working Group II: Provision of GCPs and Aerial Triangulation
3.3 Working Group III : DEM & Ortho Image Generation
3.4 Working Group IV : Data Structure
3.5 Working Group V : 2D & 3D Data capture
4. List of Participants
5. Schedule of Technical session
6. Technical Proceedings:
6.1 Inaugural session:
6.2 Lead Paper
6.3 Working Group I : Choice of Input Data
6.3.1 Presentation
6.3.2 Discussion
6.3.3 Discussion
6.4 Working Group II : Provision of GCPs & Aerial Triangulation
6.4.1 Presentation
6.4.2 Discussion
6.5 Working Group III : Generation of DEM & Ortho image
6.5.1 Presentation
6.5.2 Discussion
6.6 Working Group IV : Geospatial Data Structure
6.6.1 Presentation
6.6.2 Discussion
6.7 Working Group V : 2D & 3D data capturing
6.7.1 Presentation
6.7.2 Discussion
6.8 Comments from Industries
6.9 Panel Discussion
7. Concluding Address by Additional SG IIS&M
8. Summary
9. Recommendation




                                           4
Back ground:

EVOLUTION OF PHOTOGRAMMETRY:

         The concept of photogrammetry and its application in surface data acquisition
using terrestrial photographs was developed in the later half of 19th century. The word
Photogrammetry first appeared in the year 1893. During that period principle of
perspective geometry of photograph and stereo photogrammetry were conceptualized and
applied further in the instrumentation. Aerial Photogrammetry in its present sense was
practically started with the development of aeroplane and was made use in I World War.
By II World War Aerial Triangulation process along with variety of stereo plotting
instruments were developed to its full potential. In 1960s with the introduction of
electronic computation, Analytical solution of photogrammetry received a fresh impetus
and with the fast development of electronic computing devices the present form of Digital
Photogrammetry developed in late 1980s. By the end of the 20th century it has developed
to its full potential to replace the Analogue Photogrammetry.

PHOTOGRAMMETRY IN SURVEY OF INDIA:

         Terrestrial Photogrammetry was tried in Survey of India in early part of 20th
century and Air survey technique from aerial photographs was started during the I World
War. The Indian Air Survey Committee was formed in 1927. Subsequently various
initiatives were taken and experiments were carried out to produce 1:50,000 topographical
maps. However until II World War started topographical maps were not produced from
aerial photographs. In 1960s and 70s, the analogue stereo plotters (optical & mechanical)
were acquired in a big way and technical staffs were trained in bringing the Analogue
photgrammetric technique completely in production line. Before the Analytical phase was
implemented in the late 1980s and early 1990s the Digital Photogrammetric technology
was developed to replace the Analogue phase. With the development of technology in all
fields of geospatial technology, availability of digital input and demands of geospatial data
in digital form Survey of India has completely switched over to this technology by the
early part of this decade.

        The success of the Analogue Photogrammetry in production line is attributed to the
well planned implementation strategies. These include the capacity, defining robust
standard and establishing methodologies for each type of task. Even though digital
technology is in vogue for about a decade it has not been implemented to its full potential
in production line due to the absence of such strategies. This need a careful study and
thought on the issues involved and how it should be addressed so that there will be a
seamless transformation of the existing experience of analogue photogrammetry, which
was the forte of SOI. Indian Institute of Surveying & Mapping (IIS&M) has always been
an instrumental in such adoption of technologies through propagation and dissemination of
knowledge.




                                             5
        Keeping these factors in view, this two-day workshop on ‘Effective
implementation of Digital Photogrammetry in SOI’ was organized at IIS&M,
Hyderabad, in collaboration with Institution of Surveyors (Hyderabad Branch) during 10th
– 11th Sept. 2009.

2. OBJECTIVES OF THE WORKSHOP:

          To identify the problem areas encountered in adopting the technology.
          Sharing experiences by the officers involved in various projects being carried
           out with the involvement of industries.
          Approach towards defining standards, specification and optimum methodology
           of digital photogrammetry processes for various scale of mapping.
          Approach towards standardization of spatial data base for various scales of
           mapping.
          Improving the training modules in IIS&M and imparting training to the SOI
           officers and technical staff.
          Level of Collaboration with / support from other organizations / institutions /
           industries involved in this field.

3. FORMATION OF WORKING GROUPS:

        A limited number of participants from the department and out side who are / were
technically involved in this field and can contribute towards achieving the objectives were
identified and invited. Among the invited participants, five Working Groups were
constituted to address various issues involved with the technology. Three members from
Survey of India and one each representing from other organization / institution and
industries constituted a Working Group. Points of References considering all aspects
pertaining to the given topics were provided to each Working Group so that they deliberate
upon to specify the parameters, specification and consider the suitability for mapping on
different scales (Large & Medium scale) / specific purpose before making a presentation in
the workshop. They were specifically suggested to add on their experiences in carrying out
various tasks / projects employing Digital photgrammetry technique, in their presentation.
Selection of Organisations, Industries and Members so included in various Working
Groups was done considering the roles they are playing and also their potential and
experiences in the respective fields. The main objective of including cross section of
representatives was to gather different opinions bringing their experiences and latest
developments in this field to facilitate an open and wide range of discussion and
deliberation.




                                                6
3.1 WORKING GROUP I: CHOICE OF INPUT DATA

      Organisation                 Nominated Members                  Participation


Survey of India                1) Maj Gen S V Chinnawar,      1. Sh S Subba Rao, Director,
                                  Addl. S G (Tech) SGO           APGDC, Hyderabad
                               2) Sh M Dharmaraj, Director
                                  TN,P &ANI GDC, Chennai
                               3) Sh S Subba Rao, Director,
                                  APGDC, Hyderabad


National Remote Sensing      Dr. P S Roy, Dy. Director        Dr. Venugopal Rao
Centre(NRSC)
Hyderabad


Infotech Enterprise Ltd.,    1. Sh Hanumanth Prasad           1) Sh Hanumanth Prasad
Hyderabad                    2. K V L Varaprasad              2) K V L Varaprasad


POINTS OF REFERENCES:

       1. Possible inputs:

               (i)     Existing Aerial photos
               (ii)    Acquisition of fresh Photos
               (iii)   Stereo satellite imageries
               (iv)    ALTM / LIDAR.

        The choice of the input to be specified in terms of Scale of Mapping, Contour
Interval, Out put Product type and Purpose specific. While choosing the data following
factors were to be addressed.

       (i) Existing Aerial photos :

               Scale
               Utility / products that can be generated
               Availability of CCC, Currency and Condition
               Scan Resolution
               Image Format

       (ii) Acquisition of fresh Photos :

               Scale
               Flight Planning ( Direction, Overlap, Season etc.)
               Flying height


                                               7
                   Focal Length
                   Supplementary Data ( IMU , GPS etc )
                   Whether Frame / Digital Camera with specification

          (iii) Stereo satellite imageries & ALTM :

                   Available data
                   Specification, Geometry, associated parameters & information and
                   their implications

          2. Cost Factor
          3. Supplying Agencies
          4. How to Indent & Obtain

  3.2 Working Group II: Provision of GCPs and Aerial Triangulation

      Organisation                    Nominated Members                     Participation


Survey of India               1) Maj Gen KRMK Babaji Rao , Addl. 1) Maj Gen KRMK Babaji Rao,
                                 SG Southern Zone, Bangalore          Addl. SG Southern Zone,
                              2) Brig P S Samudra, Director. OGDC,    Bangalore
                                 Bhubaneshwar                      2) Sh P K Das, SS,
                              3) Sh P K Das, S.S.                  3) KGDC, Bangalore
                              4) K GDC, Bangalore


Anna University Chennai       No Response


Speck System Ltd.             1. Sh J N A Murthy,                  1. Sh J N A Murthy
Hyderabad                        Chief Technical Officer


Leica India                   Sh Mahesh Reddy, Technical Manager   Sh Mahesh Reddy
Hyderabad


  POINTS OF REFERENCES:

        The following factors were to be addressed in terms of input data, Scale of
  Mappings, Contour Interval, Output Products and Purpose.

          (i)     Distribution, Method of Observation and Accuracy Level of Plan & Height
                  controls and Check points
          (ii)    Co-ordinate System & Height model


                                                 8
       (iii) Guidelines for Post Pointing
       (iv) Choice of Camera / Sensor Model ( Input Specific) Distribution and
              Parameters for auto tie point generation
       (v) Parameters ( Convergence etc.) during AT
       (vi) Acceptable RMSE for GCPs & Check points(in Ground & Image coordinates
              term)
       (vii) Key factors Report Analysis
       (viii) Report Recording Methods

3.3 Working Group III : DEM & Ortho Image Generation

      Organisation                    Nominated Members                   Participation


Survey of India              1)   Maj Gen S V Ravi,               1) Maj Gen S V Ravi,
                             2)   Addl. SG M&G GDC, Hyderabad     2) Addl. SG M&G GDC,
                             3)   Sh P K Ganguly,                    Hyderabad
                             4)   Director. M & A GDC, Shillong   3) Sh P K Ganguly,
                             5)   Sh S K Sinha,                   4) Director. M & A GDC,
                             6)   Dy. Director, DSSDI,               Shillong
                             7)   New Delhi                       5) Sh S K Sinha represented
                                                                     Working Group V


ADRIN                        Dr. P V Radhadevi,                   Dr. P V Radhadevi
Hyderabad                    Head DM & MD


COWI                         Sh Prashant Srivastava,              Sh Prashant Srivastava
Gurgaon                      Sr. Manager Photogrammetry


POINTS OF REFERENCES:

      The following factors were to be addressed in terms of input data, Scale of
Mappings, Contour Interval, Out put Products and Purpose :-

       (i)     Methodology for generating DEM / DSM
       (ii)    Grid Size
       (iii)   Error Evaluation Process & Methodology
       (iv)    Acceptance Accuracy
       (v)     Methodology for generating Ortho Image
       (vi)    Resampling and output pixel size
       (vii)   Positional accuracy evaluation




                                              9
3.4 Working Group IV : Data Structure

  Organisation             Nominated Members                       Participation

Survey of India    1) Maj Gen P N Koul, Addl. S G       1)   Maj Gen P N Koul,
                      (HRD) SGO                         2)   Addl. S G (HRD) SGO
                   2) Sh P V Rajshekhar,                3)   Sh P V Rajshekhar,
                   3) Director. J & K GDC, Jammu        4)   Director. J & K GDC, Jammu
                   4) Sh R N Nanda, S.S., GIS&RS,       5)   Sh R N Nanda, S.S.GIS&RS,
                   5) Hyderabad                         6)   Hyderabad


JNTU               No Response
Hyderabad

                                                        Sh. Dhiraj Goel
ESRI India         Sh. Dhiraj Goel, Project Manager
New Delhi          NIIT GIS


POINTS OF REFERENCES:

       The following factors were to be addressed in terms of, Scale of Mappings :-

       (i) Geometric and Attribute structure Feature wise
       (ii) Feature Organisation
       (iii) Data Filtering and Generalisation

3.5 Working Group V : 2D & 3D Data capture

 Organisation / Industry         Nominated Members                      Participation

Survey of India              1 Brig Girish Kumar,               1. Sh S K Sinha,
                             Project Director, DSSDI, New       Dy. Director, DSSDI,
                             Delhi                               New Delhi
                             2. Sh B K Mahapatra,               ( Sh S K Sinha represented
                             Superintending Surveyor,           DSSDI as Maj Gen Girish
                             IIS&M                              Kumar could not be present
                             3.Sh T P Mallick,                  for health reason)
                             Superintending Surveyor,           2. Sh B K Mahapatra,
                              T, P A& N GDC,                    Superintending Surveyor,
                             Chennai                            IIS&M

IIRS,                        Ms. Shefali Agrawal                Ms. Shefali Agrawal
Dehra Dun                    In charge Photogrammetry
                             Division

RSI,                         1. Sh Gangadhar Rao                1. Dr. Malleshwar
Hyderabad                    2. Dr. Malleshwar                  2. Sh Srinivas Reddy



                                            10
POINTS OF REFERENCES:

      The following factors were to be addressed in terms of input data, Scale of
Mappings, Contour Interval, Output Products, Data Structure and Purpose :-

      (i) Choice of 2D / 3D
      (ii) Capturing methodology
      (iii) Optimum Zoom level & guide line for data entry
      (iv) Contour generation
      (v) Method of scrutiny ( both positional & attribute)
      (vi) Error Evaluation & Accuracy acceptance level
      (vii) Out turn level
      (viii) Record maintaining method

4. LIST OF PARTICIPANTS

      Institution of Surveyors ( Hyderabad branch):

              1)    Sh T C Neelakantan
              2)    Sh G S Kumar
              3)    Sh N K Agrawal
              4)    Sh P Upendra

      Survey of India (Out Station)

              1)     Maj Gen Manoj Tayal, Addl. Surveyor General
              2)     Sh Nagendra Prasad, Addl. Surveyor General
              3)     Maj Gen KRMK Babaji Rao, Addl. Surveyor General
              4)     Maj Gen P N Koul, Addl. Surveyor General
              5)     Maj Gen S K Pathak, Addl. Surveyor General
              6)     Maj Gen P K Vachher, Addl. Surveyor General
              7)     Sh R M Tripathy, Director
              8)     Sh P K Ganguly, Director
              9)     Sh Bhagirathi Mahapatra, Director
              10)    Sh P V Rajashekhar, Director
              11)    Sh S K Singh, Director
              12)    Sh R. A. Nirvikar, Director
              13)    Sh S K Sinha, Dy. Director
              14)    Sh P K Das, Superintending Surveyor

      Survey of India (Hyderabad Station)

              1)     Maj Gen (Dr.) B Nagarajan, Addl. Surveyor General
              2)     Maj Gen V R Mahendra, Addl. Surveyor General
              3)     Maj Gen S Ravi, Addl. Surveyor General



                                         11
        4)     Maj Gen V S Deo, Addl. Surveyor General
        5)     Sh S Subba Rao, Director
        6)     Sh Ashok Prim, DSG
        7)     Sh J K Rath, DSG
        8)     Sh R N Nahak, Director
        9)     Sh P V Srinivas, DSG
        10)    Sh A K Das, SS
        11)    Sh B K Mahapatra
        12)    Sh B N Das, SS
        13)    Sh B Mahanta, SS
        14)    Sh A Purnachandra, SS
        15)    Sh R N Nanda, SS
        16)    Sh G C Nayak, SS

Participants from other Organisations

        1) Dr. Venugopal Rao, NRSC
        2) Dr. P V Radhadevi, ADRIN
        3) Ms. Shefali Agrawal, IIRS

Participants from Industries

        1)    Sh Hanumanth Prasad, Infotech Enterprises Ltd.
        2)    K V L Varaprasad, Infotech Enterprises Ltd.
        3)    Sh J N A Murthy, Speck Systems Ltd
        4)    Sh Mahesh Reddy, Leica India
        5)    Sh Prashant Srivastava, COWI
        6)    Sh. Dhiraj Goel, NIIT GIS
        7)    Dr. Malleshwar, RSI
        8)    Sh Srinivas Reddy, RSI




                                    12
5. Schedule of Technical session

                                       10th September 2009

09:00 – 09:30 hr   Registration

                   Inauguration by Maj Gen Manoj Tayal, Additional Surveyor General , Central Zone
09:30 – 10:30 hr   Lead paper by Maj Gen (Dr.) B Nagarajan, Addl. SG, IIS&M, Hyderabad –
                   Setting the agenda for the workshop
10:30 – 11:00 hr   Tea Break

11:00 – 12:00 hr   Presentation by Working Group I : Choice of Input Data
                   Lead by Sh. S Subba Rao , Director, APGDC, Hyderabad
                   Discussion
12:00 – 13:00 hr   Presentation by Working Group II : Provision of GCPs & Aerial Triangulation
                   Lead by Maj Gen KRMK Babaji Rao, Addl. SG , Southern Zone, Bangalore
                   Discussion
13:00 – 14:00 hr   Lunch Break

14:00– 15:00 hr    Presentation by Working Group III : Generation of DEM & Ortho image
                   Lead by Maj Gen S Ravi, Addl. SG, M&G GDC, Hyderabad wing
                   Discussion
15:00 – 16:00 hr   Presentation by Working Group IV : Geospatial Data Structure
                   Lead by Maj Gen P N Koul, Addl. SG, SGO, Dehradun
                   Discussion
16:00– 16:15 hr    Tea Break

16:15 – 17:30 hr   Presentation by Working Group V : 2D & 3D data capturing
                   Lead by Sh. S K Sinha, Deputy Director, DSSDI, New Delhi
                   Discussion
                                       11th September 2009

                   Training Modules
                   Lead by Maj Gen (Dr.) B Nagarajan, Addl. SG, IIS&M, Hyderabad
                   Discussion
                   Contribution / Comments from Industries

                   Tea Break

                   Panel Discussion & Deliberation on working groups Reports
                   Recommendation & Closing ceremony followed by Lunch




                                            13
6. TECHNICAL PROCEEDINGS:

6.1 Inaugural session:

        Sh Ashok Prim, DSG, IIS&M welcomed the Workshop participants and invitees to
the Workshop and read out the message of best wishes and compliments for the successful
conduct of the Workshop sent by Maj Gen R S Tanwar , Surveyor General of India, who
could not make himself available on the occasion owing to some other engagement.
        The Workshop was inaugurated by Maj Gen Manoj Tayal, Addl. Surveyor General
Central Zone. In his inaugural address he re-emphasized the role of Photgrammetry in
Surveying and Mapping and analysed relevance of the Workshop on „ Effective
Implementation of Digital Photogrammetry in Survey of India‟ in the present context when
the technology is in vogue and considering the volume of tasks ahead. He lauded that the
conduct of Workshop is being conducted at appropriate time as the department, being the
premier Mapping Agency of the Govt. of India, could not afford to delay further in
effectively implementing the technology. He hoped that all the Working Group would
come up with recommendations after due deliberation, which would be helpful to the
department and wished all success of the Workshop.

6.2 Lead Paper

         Maj Gen (Dr.) B Nagarajan, Additional Surveyor General IIS&M set the tone of
the Workshop through presenting the lead paper where he set the agenda for the Worshop.
In his address he dealt on the status and stages of evolution of Photogrammetry through
various phases of Analogue, Analytical and Digital. He gave an account how in the
department the Analogue Photgrammetry was implemented systematically in a planned
manner during 1970s – 1980s through sequentially developing infrastructure and building
capacity, which was the reason why the technology could be implemented successfully and
was the forte of SOI. In contrary the same was absent while adopting the Analytical and
Digital technologies. It is a matter of concern that the technology is growing fast and the
department is not able to keep pace with it, while the industries are ahead in adopting to its
full potential as far as data capturing is concerned. He expressed his conviction that SOI is
still in strong footing as far as theoretical foundation is concerned. Only it has to be
implemented in its production line in a systematic and time bound manner where
technical staff will get sufficient exposure and gain practical experience.

        He explained the background and thought process involved that influence in
organizing this Workshop, identifying the topics and constitution of five Working Groups.
He appealed to all to deliberate upon all aspects of the technology wit an open mind in
stead of a set closed mind, so that the department can fulfill its mandate in providing
geospatial data in time as per the requirements of the Government of India and other
agencies. He impressed upon all the participants and Working Groups‟ members to
consider the issues of cost effectiveness, purpose specific scale of the products and extent
of out sourcing while deliberating on various topics. He also hoped that the outcome of
the workshop would provide a direction to reorient the course curriculum of IIS&M.




                                             14
        He concluded with the remark that the National Developments and Security must
be the prime consideration factor, which should always be kept in mind while adopting any
technology.

6.3     Working Group I : Choice of Input Data

        Chair                :        Maj Gen (Dr.) B Nagarajan
        Presentation Lead by :        Sh S Subba Rao

6.3.1   Presentation

1.     Sh S Subba Rao made a brief presentation with an opening remark that the points
presented were to trigger the discussion in detail. The main points of his presentation
were :

(i) Possible options of input data:
        (a) High Resolution Satellite Imagery (Stereo). The specification of sensors of
                 Geo Eye 1, World View 1 and Cartosat 1 satellites that are available today.
          (b)   Existing Aerial Photography (Hard copy) Presented the specification of
                 Colour Photogrammetric scanner DeltaScan
          (c)   Acquisition of Digital Photograph Presented the specification of Digital
                 camera (Airborne scanner) 3-DAS-1
          (d)   ALTM / LiDAR.


(ii) Suitability
       (a) Aerial Photograph

       It can provide plan and height accuracy up to 0.25 m and hence suitable for large
scale utility, urban, infrastructure development and cadastral mapping. Thus one time
investment can be recovered from such mapping.

       (b) High Resolution satellite Imagery
       The available high resolution satellite stereo imagery can provide adequate plan
accuracy, but in the context of height accuracy it is not suitable foe mapping on scale
1:10,000 or larger.

(iii) Cost Effectiveness

       Cost of generating 1:2,000 scale maps from 1:10,000 scale aerial photography is
Rs.30,000 per sq. km with an accuracy of 25 cm in both plan & height. The cost of
generating it from high resolution satellite imagery is Rs 60,000 per sq. km with out
achieving the same accuracy.



                                             15
(iv) Case Study

        The process involved in generating 1:2,000 data base for NUIS project work in
respect of Nalgonda town from 1:10,000 analogue aerial photographs, was presented.

2. Presentation by representative from Infotech Enterprises Ltd.

      Sh Hanumanth Prasad representing Infotech Enterprises Ltd., as a participant in
the Working Group I enumerated categorically the choice of data for different purpose.

       (a) Comparison

        He compared the Aerial Photography with Satellite Imagery in the context of urban
mapping, specially in respect of imaging building. Considering the nature of the sensor and
height of the platform Aerial Photographs will provide sharper image in comparison to
Satellite Images.

       (b) Requirements

       He catagorised the requirements in to three categories and suggested choice of
input data for each category as follows :

        (i)   Urban Utility Mapping – Aerial Digital Image
        (ii) Topographic mapping – Satellite Stereo image
        (iii) Thematic mapping – Satellite Mono Image

       He also suggested that a hybrid data like high resolution Ortho satellite image,
generated using DEM from Cartosat -1 data, can be used for mapping on scale 1:15,000.

3. Presentation by representative from NRSC

       Dr. Venugopal Rao representing NRSC briefly discussed the advantages of using
Cartosat-1 stereo images, basing on his experience.

       (a) Comparison
       Generating a 1:10,000 maps from Cartosat – 1 images is cost effective and less
time consuming compared to,generating it from Aerial Photographs. Carrying out Aerial
Triangulation (Block Triangulation) using available RPC is easier in case of Cartosat – 1
images. With about 5 points,Block Triangulation can be performed for a stretch of 100 sq
km area.

       (b) Data availability

       Cartosat – 1 stereo images are available for about 96% of the country.




                                            16
        (c) Experiences

       1,50,000 sq km area in costal region has been mapped from Cartosat – 1 stereo
images.

        (d) Accuracies achievable

        3 – 4 meters in both plan and height is achievable

        (e) Advantage over foreign satellite data products

         (i) Satellite like IKONOS and Quick bird provide stereo data with different
               geometry which necessitate more numbers of images for an area in
               comparison to      Cartosat – 1.
         (ii) Cartosat -1 is more stable then other available satellites including Cartosat -
               2.
         (iii) Block adjustment is easier in case of Cartosat – 1 images.

       Dr. Venugopal Rao while appreciating the accuracy standard of SOI topographic
maps on scale 1:50,000 & 1:25,000, opined that as per the requirement, the laid down
accuracy specifications for various scales may be compromised so that demands of large
scale maps can be fulfilled and data can be provided to the agencies in time, timely using
available satellite data like Cartosat – 1.

6.3.2   Discussion

       In response Dr. Venugopal Rao advocated for Cartosat – 1 images in place of
Foreign satellite data as the former is more stable and cover much more area in a single
scene.

       Maj Gen Tavinder Paul while appreciating the efforts of NRSC joined the issues
and stated that SOI the National Mapping Agency can not compromise on the accuracy
standard of maps which form permanent record for various purpose unlike project specific
maps which are temporary in nature and purpose specific.

         The chair Maj Gen (Dr.) B Nagarajan moderated the discussion and emphasized
that all should be open minded to listen to the views of other organizations and industries
to explore the possibilities for providing a solution.

         (i)        Maj Gen Tavinder Paul questioned regarding the proposed input data for
                    1:10,000 scale mapping in terms of cost and requirements of control points.




                                                  17
       In response Dr. Venugopal Rao advocated for Cartosat – 1 images in place
       of Foreign satellite data as the former is more stable and cover much more
       area in a single scene.

       Maj Gen Tavinder Paul while appreciating the efforts of NRSC joined the
       issues and stated that SOI the National Mapping Agency can not
       compromise on the accuracy standard of maps which form permanent
       record for various purpose unlike project specific maps which are temporary
       in nature and purpose specific.

       The chair Maj Gen (Dr.) B Nagarajan moderated the discussion and
       emphasized that all should be open minded to listen to the views of other
       organizations and industries to explore the possibilities for providing a
       solution.

(ii)   Sh T C Neelakanthan stated that accuracies of any map data is subject to
       ground verification, which forms an inseparable part of mapping.

(iii) Maj Gen P N Koul suggested for a group to carry out pilot projects on
      different terrain.

(iv)   Sh S K Sinha asked some clarifications about questioned the geometry of
       satellite sensors, which play a crucial role in mapping.

(v)    Sh G S Kumar intervened in the discussion and said that he could find any
       claim of large scale mapping with authority using satellite imageries, in
       Internet.

(vi)   Sh P V Rajashekhar viewed that apart from other aspects, cost benefit vis-a-
       vis time taken to provide data and simplicity of technology must be
       weighted. In support of his view he cited an example of increase in tax
       collection by making large scale urban maps available in time, in the state
       of Karnataka.

(vii) Dr. P V Radhadevi and Sh JNA Murthy advocated for use of Cartosat – 1
      stereo data as it is more stable and covers more area in comparison to other
      data. On this issue the chair Maj Gen (Dr.) B Nagarajan supported saying
      foreign satellite data are much more costly affair as number of image pairs
      increases for the same area.

(viii) Dr P V radhadevi also supported for hybrid data to satisfy the requirements.

(ix) Ms. Shefali Agrawal suggested to try with hybrid data specific to various
     types of terrain.




                                    18
6.3.3   Conclusion

        Maj Gen (Dr.) B Nagarajan, chair of the session thanked all the Working Group
members and participants for lively debate and summerised the session before closing it.
He viewed that we should not be confined to the one mind set and open to the development
going in the industries so that SOI can serve the requirements for national development,
security and disaster management.

6.4 Working Group II : Provision of GCPs & Aerial Triangulation

    Chair                :    Maj Gen Tavinder Paul
    Lead Presentation by :    Maj Gen KRMK Babaji Rao

6.4.1   Presentation

        Maj Gen KRMK Babaji Rao made the presentation supported with results of Aerial
Triangulation (AT) carried out for various large scale urban mapping using Analogue
aerial photographs. The main points of presentation were as under :

(i) Basics

        (a) Maj Gen KRMK Babaji Rao explained the basics of relationship between
            image & Ground Coordinate system in a perspective geometry of Aerial
            Photographs. Task of Aerial Triangulation is to find the Parameters to establish
            these relation ship for further photogrammetric process.

        (b) He made a comparison between the AT process in analogue environment
            (Independent Model Triangulation & Block adjustment using PATM 43) and in
            Digital environment where Exterior Orientation Parameters (X0, Y0 , Z0 , Κ,Φ,
            ω) of each exposure station are found by measuring some image space and
            corresponding Ground space coordinates; the mathematical model remaining
            the same in both.

(ii) Distribution of Control Points

         (a)   As the mathematical model remains the same, the calculation of Bridging
               distance for both plan and height control remains the same as required in
               analogue technique for the same kind of input and out put requirements.

         (b)   Plan control is required on the periphery of the block only at the specified
               bridging distance (say 4-5 model apart) . The control will be located on at
               least one model falling completely outside the block with a few scattered
               points inside the block to serve as check points.




                                             19
        (c)   Height control points are provided in chains (in the upper and lower internal
              over laps) at right angles to the direction of flight at the specified bridging
              distance and at the beginning and end of the strip




(iii) Additional Information

        (a)   Aerial photographs accompanied with GPS data provide X0, Y0 , Z0 of the
              exposure station only.
        (b)   GPS data with Inertial Motion Unit (IMU) provide X0, Y0 , Z0 , Κ,Φ, ω of
              the exposure station.
        (c)   With the availability of these data, the number of control points required
              will be reduced. However in Indian data set these data are not available at
              present, hence not tried in any project work.




                                            20
(iv) Case Study

        Maj Gen KRMK Babaji Rao presented the experiences of carrying out Aerial
Triangulation for large scale urban mapping in respect of Chennai, Kolar, Raichur, Ballary,
Bidar and Palakkad using analogue aerial photographs with out GPS / IMU data. The
scale of photography is 1:6,000 for Chennai and !:8,000 for other towns. The distribution
of control points, where optimum solution were found, with block diagram and obtained
accuracy is enclosed as Annexure I. Based on these experiences the followings are
recommended.

(a) GCPs

      Two GCPs each in the over lap areas in first and last models of first and last strips.
       GCPs in lateral overlap areas to have more than four rays (GCPs on > four images).




      In all other strips one GCP in lateral overlap at the beginning and at the end.
      All models adjoining gaps or large water bodies to have one GCP.




      One GCP every fifth model/ photo. All GCPs are to be as far as possible as Plan
       and Height. However, they have no effect on the plan accuracy of the block.

      He contested some of the claims of industries that it can be done with fewer control
       points on the premises that Indian data does not have IMU information and
       accuracy levels of the process need to be established. For that reason post
       processing field verification is required. He put in favor of preprocessing provision
       of GCPs so that sufficient numbers of it can be used as check points.
      He suggested that the minimum number of control points should not be less than
       less than 10% of the number of exposures it takes to cover the project area at the
       prescribed photo scale. Out of them use as many of the points as possible for check
       points. Additionally, whenever possible, locate GCPs that lie on multiple images
       and around the edges of the block.




                                             21
      The post pointing should be done on the copy of sufficiently enlarged print of the
       photograph to be used in the processing. It should be clearly identifiable in
       photograph up to pixel level and on the ground. Other image, may be latest one,
       must not be used as the photographs may be of older vintage where the
       detail/object on the ground may not be appearing on the photograph. It is
       suggested that the post pointing is to be carried out on soft copy of the photographs
       duly verifying the identifiability of the point on other photographs.

      The sketch which was in vogue earlier prepared on 1 PHOT form may be replaced
       by this enlarged phtotograph and supported with description.

(b) Check Points

        (i)     At least 10% of the no. of GCPs planned, be provided as check points.

        (ii)    Check points should be provided in the centre of the block, between the
                GCPs, and in the strip (lateral) overlaps.

(c) Tiepoints

        (i)     It has been experienced that checking and correcting tie points generated by
                auto tie point generation techniques is much more time consuming and
                hence needs more patience, expertise and more error analysis capability.
                Thus manual measurements are preferred in such input data.

        (ii)    Distribution should be min 3 columns X 3 rows in each photo. Optimum
                recommended is 3 columns X 5 rows in each photo. For tough areas 5
                columns X 5 rows in each photo.

        (iii) Central tie point to be on min of three images.

        (iv)    Upper and lower points to lie within the lateral overlaps of the strip. Each
                tie point to be on six photos as far as possible. If these tie points are good
                then the stitching of the block is going to be perfect.

        (v)     During adjustment, as GCPs are given higher weightage, error in GCPs is
                distributed to tie points. If tie points are checked and perfect then one can
                easily sort out erroneous GCPs.

(d) Parameters

        (i)     Standard Deviation of the Image measurements: 1/3 of the pixel
        (ii)    Standard Deviation of the Controls Points: Fixed / as per the GPS survey
        (iii) No of Iterations: 10
        (iv)    Convergence : 0.00010 meters / project specific unit


                                              22
        (v)      Blunder detection can be time saving/robust; blunder check can be used for
                 identifying blunder points.
(e) Acceptance Tolerance

         (i)     RMSE for GCPs & Check points(in Ground & Image coordinates term)
         (ii) Total Image unit – weight RMSE (Sigma 0) = 1/2 Pixel
         (iii) Root Mean Square Error (RMS) of residuals in X,Y & Z at the Full controls
                  < 1/2 Pixel
         (iv) Root Mean Square Error (RMS) of residuals in X,Y & Z at the check Points
                  < 1/2 Pixel
         (v)     RMS value of residuals in image coordinates < 1/2 pixel

(f) Key factors for Report Analysis

           Criteria to be checked after the block adjustment are:

               (i)      Blunders
               (ii)     Sigma 0 of the solution
               (iii)    RMS of residuals of control points
               (iv)     RMS of residuals of Tie points
               (v)      Systematic effects at residuals of control points
               (vi)     GPS observations
               (vii)    IMU attitude observations
               (viii)   Quality of adjusted terrain points and projection centers
               (ix)     Error ellipses
               (x)      Reliability of points and projection centers
               (xi)     Reliability „rectangles‟
               (xii)    Tie Coverage to find “holes” in the block

(g) History Sheet

       Along with other records the followings should be included in the history sheet of
Aerial Triangulation.

               (i) Aerial Triangulation Summary
               (ii) Exterior Orientation Parameters results
               (iii) Control point , Check point coordinates and accuracy estimates




                                                  23
(h) Rethinking on Accuracy

       In the past the output from photogrammetry was a hard copy map only. Therefore
the requirement for accuracy consideration was plotting error. Accordingly accuracy of
control was set higher than half of the Plottable Error on mapping scale.

   Example:    1:10,000 mapping accuracy of control > 1.25 meter,
               1: 5,000 mapping accuracy of control > 0.625 meter

       In the present situation the output from photogrammetry or any survey is the digital
topographic data base, which will be input again for GIS, for compilation of maps at
various scales and production of hard copy maps after converting into cartographic data
base. Further during Aerial Triangulation process Root Mean Square Error (RMS) of
residuals in X,Y & Z at the Full controls is set at < 1/2 Pixel.

       Therefore in digital environment we should be thinking in terms of pixels and
not in terms of Plottable Error.
       Thus GCP accuracy should be > ½ GSD (Ground Sampling Distance) Half of
the pixel size in ground term.
       However it has to be appreciated that no out put can be better than the input
data in respect of accuracy, quality of interpretation.

(i) Presentation by Representative of Speck Systems Ltd.

         Sh J N A Murthy Presented experiences of various projects undertaken by Speck
Systems Ltd. Specially he mentioned about the Cadastral mapping project of Nizamabad
district where Digital Photogrammetry has been used. The area coverage is 8000 sq km.
Based on this he stated that;

            Cost of Aerial Photographs is only 25% of the project cost.
            GCPs at 5 photos apart gives the optimum solution
            According to him some of the GCPs should be prepointed by signal for
             improving the Aerial Triangulation accuracy.

(j) GCPs for satellite stereo images

        Dr Venugopal Rao stated that satellite stereo imagery like Cartosat-1 would need
minimum number of control points. According to him 4 GCPs in a block may be sufficient
for getting the required accuracy. Sh J N A Murthy suggested that 9 control points in three
bands will solve in all types of area. The chair Maj Gen Tavinder Paul intervened and
invited Sh G C Nayak to state the experiences of the project work carried out at IIS&M. In
response based on the evaluation at three sites Sh G C Nayak stated that 5 control points, 4
in corners and 1 in middle, were sufficient to get the optimum result in one scene of
Cartosat -1 stereo imagery with RPC. Evaluation of multi stereo pairs in a block is in
progress, at IISM.




                                            24
6.4.2   Discussion

        (a) Sh P K Ganguly raised the questions of redundancy for the solution and Level
of accuracy checking.
        In response Maj Gen Babaji Rao stated that the suggested pattern of GCPs & Tie
points introduce the sufficient redundancy and recommended provision of 10% of number
of GCPs as check points.
        He claimed that in Chennai project with this pattern accuracy through ground
truthing was found to be 99% accurate and ground measurement found to be agreed within
7 centimeters. Sh B K Mahapatra supported the claim of accuracy citing the case of
Ahemadabad project.

      (b) Regarding Tie point Generation Maj Gen Babaji Rao strongly advocated for
Manual Tie point Generation as blunder checking in auto tie points is a tedious & time
consuming affair.

        Dr. Venugopal stated that auto tie points generation is up to 60% accurate
irrespective of S / W used.

       In response to this Sh Hanumanth Prasad stated that Bingo S/W can give 90%
accurate.

        The same claim was contested by Sh Dhiraj Goel by staing that even though RMSE
value may be showing less but in stereoscopic measurement the agreement may not be that
good and instead claimed that Socket set provide better result. He further suggested that
error in tie points can be removed in phase wise manner rather than generating manually
which is laborious and costly affair especially in a large block.

       Maj Gen Babaji Rao stated that his suggestions is based on the context of Indian
condition where quality of input is not that superior and GPS / IMU data are not available.

        Sh Mahesh Reddy responded that with the availibilty of GPS / IMU data LPS can
also provide the the same result.

       The chairman and Maj Gen (Dr.) Nagarajan intervened to say that discussion may
not be S/W specific rather on general Solution.

        Sh P K Das added that high percentage of overlap (more than 60%) will give better
result of auto tie point generation in hilly area.

        Sh P V Rajashekhar suggested that model wise tie point generation can provide
better result as pari pasu checking can be done.

       iii. Sh Hanumatth Rao , Sh Prashant Srivastava and other from industries told that
industry are way ahead in the technology as they have gained experiences from various
types of jobs.



                                            25
6.5 Working Group III : Generation of DEM & Ortho image

    Chair                     :       Maj Gen P N Koul
    Presentation Lead by      :       Maj Gen S Ravi

6.5.1   Presentation

       Maj Gen Ravi made the presentation with the opening remark that it was based on
       the
input provided by the Working Group members, specially the project experiences of Sh
Prashant Srivastava and experiences of Dr. P V Radhadevi with Cartosat 1 imagery,
contributed immencely to the presentation.

(i) Fundamentals

        He defined the ortho rectification which is carried out to rectify the relief and tilt
displacement present in a perspective photographs to obtain a orthophoto. However it is
stated that orthophotos may not be considered as a substitute for a precise line map since
cultural features with a vertical elevation are not as accurate or discernable on an
orthophotograph as a planimetric map produced in a stereoplotter. Orthophotographs can
be used with caution as a substitute for a planimetric line map for certain smaller scale,
nondesign applications.

       The following recommended relationship between the map plot scale and
approximate pixel size of orthophoto is presented.

           Final Map Plot Scale               Approximate Ground Pixel
                                                Resolution Required

               1:500                                 0.0625 m
               1:1,000                               0.125 m
               1:1,500                               0.250 m
               1:2,000                               0.375 m
               1:2500                                0.5 m

(ii) Requirement

(a) Workstation

        1. Hardware

       A basic component of an orthophoto workstation is the central processor. The
processor must be capable of interactively processing elevation models and gray-tone
images. Primary requirements are high processing (RAM and ROM) speed, large memory
and mass storage, high-speed graphics processor, and input and output devices.



                                             26
        Other supplementary devices may be necessary for communication, data transfer,
network links, film writing, screen digitizing, high-resolution metric scanners, and various
printing devices. These devices must be compatible with the accuracy of the final
orthophoto products. For example, the scanner must be capable of producing the required
scanned image necessary for the final orthopohoto accuracy.

       2. Software

            (a) Operating system for the workstation.

            (b) Application software required for the processing image matrices and the
                creation of elevation models.

            (c) Orthophoto package used to rectify the image pixel matrix with the
                elevation model in the final step to create the orthogonal image

(iii) Ortho Photos generation route

            (a)   Photo scanning(diapositives or film)
            (b)   Collect digital terrain model
            (c)   Merge scanned images and digital elevation models
            (d)   Radiometric correction and tiling
            (e)   Data formatting and compression

(iv) General Production Procedures

       (a) Image enhancement
       (b) Image rectification

             (i) Simple rectification ( rubber-sheeting)
            (ii) Differential rectification

       (c) Image scan.
       (d) Transport image data
       (e) Orientation

            (i) Interior ("inner") orientation
           (ii) Exterior ("absolute")orientation

       (f) Produce elevation model
       (g) Geometric transformation.
       (h) Mosaicing




                                             27
(v) DTM, DEM & DSM(TIN)

      (a) Mass points and Breaklines together are considered as a digital terrain model
          (DTM).
      (b) An end user may require only mass points collected on an evenly spaced grid
          (every 10 m). This type of elevation model is considered a digital elevation
          model (DEM).
      (c) A DTM is often imported into software that will generate a Triangulated
          Irregular Network (TIN) model. A TIN is often referred to as a surface model.

(vi) DEMProduction Procedures

      (a) Utilizing mass points and break lines to denote the major changes in
          topography (DTM) can be generated in the stereo plotting device. The DTM
          can then be used to produce a Triangulated Irregular Network (TIN) of the
          topographic surface. Software can then generate an accurate grid of points over
          the TIN. The grid is at a specified interval posting. The closer the spacing the
          more accurate the ground character.

      (b) A grid of points along with mass points and break lines denoting abrupt
          changes in topography can be read and captured from direct readings on stereo
          images in a stereo plotter or softcopy workstation. This method would
          theoretically provide the most accurate elevation model. However, the time and
          subsequent cost difference over other less costly method would increase. Each
          point must be read by the stereo operator. In most cases this method of
          elevation data collection is not warranted.

      (c) A software process known as autocorrelation may be employed to establish a
          grid of mass points at a specified interval. This method uses software that
          allows the computer to automatically collect points at a specified grid over a
          stereo image. This method is very quick. However, editing must be
          accomplished to ensure that the vertical location of all points is on the earth
          surface at the specified location. The initial vertical location can sometimes be
          the top of a tree or building. Software and stereo operator editing must then be
          accomplished to edit these type of points and establish their vertical location on
          the earth surface. The mass points and break lines are then captured in a manner
          similar to that described above. This method may save time and cost in certain
          areas (i.e., areas with minimal vegetation and planimetric features).

(vii) Enlargement Factor

      (a) The enlargement from the aerial photo scale to the final orthophoto map is
          critical in the orthophoto process. This may vary between 4 to 10 times the
          photo negative scale.




                                           28
       (b) The enlargement factor is dependent upon several following items and
           conditions that may be unique to the equipment and or the project area.

       (c) Type of terrain and vegetation the images are to cover, final scale of the
           orthophotos, the elevation model accuracy.

(viii) Quality of Ortho Photo

       The quality of the orthophotos depends on the following factors :-

               (a) Aerial photography
                       Quality of film
                       Camera & lens assembly
                       Forward motion compensation
                       Quality of chemicals etc used for film processing
               (b) Quality of film scanner
               (c) Magnification
               (d) Scan resolution
                       Microns, Dots per inch (dpi), Ground Sample Distance
               (e) Method of generation of DEM
               (f) The software used for Ortho rectification

(ix) Building Modeler

        The algorithm for rectifying a perspective projected roof top of a building and its
limitation was presented with illustration and some results.

       It is concluded from this explanation that Ortho Rectified is not true ortho.

(x) Presentation by Representative from COWI India Pvt. Ltd.

        Sh Prashant Srivastava Presented a case study through project execution report of
generating ortho photos for Myambat project carried out by them. There he emphasized
the strength of INFO S/W and methodology adopted. The project area is approximately 1700
hectares covered by 33 images at 0.25m GSD captured by DMC Camera. From this, 19 images
were used for ortho production of the Area of Interest with output GSD 0.25m.

(a) Production methology

       (i)    Enhancement of the raw images
       (ii)   Ortho image generation with out put 0.25m GSD by marking the ortho areas and
              using bilinear resampling.




                                             29
       (iii) A single mosaic was generated using Feathering option Inverse Distance with
             Feather Size 80 pixels on each side without “Blending of Output Area
             Borders after checking of ortho edges, automatic radiometric adjustment and
             manual editing of automatic generated seam lines.

       (iv) Creation of tile grid

       (v)   The created mosaic was compressed to 5:1 ratio and inspected for its
             completeness, positional accuracy by overlaying vector layers and radiometric
             discrepancies. For fixing non-radiometric discrepancies, the DTM for that area
             was investigated in Stereo and on being found correct; a localized generation
             of the ortho was done for those specific areas and again adjusted.

(b) Final Quality Check

                 The entire area was completely checked for area coverage along the boundary
of Area of interest at a higher zoom level.

(xi) Presentation by Representative of ADRIN

       Dr. P V Radhadevi made the presentation on the potential of satellite Stereo
imagery for large scale mapping with some satellite specific information result of a pilot
study.

(a) High Resolution Indian Remote Sensing Optical Satellites

               (i) IRS-1C/1D 5.8m across-track stereo
               (ii) TES :1m resolution. Step and Stare technology
               (iii) IRS-P6 : 5.8m. (MX)L4, 23m (MX)-L3, 58m (MX) AWiFS- fixed
                      cameras
               (iv) Cartosat-1 :2.5 m PAN Along-track stereo –fixed cameras.
               (v) Cartosat-2 :0.8m.Pitch steering, Super resolution imaging capability,
                      Variable integration time option
               (vi) Cartosat-2A 0.8 m
               (vii) Chandrayaan – 1 – TMC camera 5m Stereo triplets.

(b) Important Factors for Mapping

               (i)  GSD of the image
               (ii) Viewing geometry/agility of the satellite
               (iii)On-orbit Calibration of the sensors
               (iv) Sensor model used for ground to image and image to ground
                    transformation
               (v) Rigorous matching
               (vi) Ground control points




                                            30
(c) Rigorous Sensor Model developed at ADRIN

        The model is developed based on the viewing geometry of the satellite and
combining the principles of photogrammetric collinearity equations. It is flexible and
adapted to a wide class of linear array sensors like SPOT-1, IRS-1C/1D, TES, IRS-P6 and
Cartosat-1, SPOT-5, Cartosat-2/2A etc with minimum number of GCP requirement. It is
part of the operational product generation chain.

        She explained how the Transformation matrix is applied –
        From CCD to Pay Load to Body to Orbit to Inertial transformation matrix.
        Through the rotation matrix, any pixel in any camera can be projected on the
ground and through inverse rotation matrix, this ground location can be re-projected to any
other camera by using the corresponding parameters.

(d) Products

        She showed slides of the following products
                i. IRS 1 D ortho image, digitized vector data from it and overlaid over
                Topo map.
                ii. images from AWiFS, LISS III & LISS IV overlaid over Topo map.
                iii. Orthoimage and Contour of Everest generated from Cartosat 1 draped
                over DEM
                iv. 3 D view of a part of moon captured by Chandrayaan 1 Terrain Mapping
                Camera
        e. Pilot Study
        She illustrated the comparison results from Cartosat 1 , Cartosat 2 and Worldview 1
in pilot study to draw the following conclusion.
                i. Cartosat-1 DEM, geometric accuracy and capability for topographic
                feature capture are good enough for making 1:10000 scale of maps.

               ii. Geometric accuracy and feature detectability of Cartosat-2/2A indicate
               that it is capable of making 1:7000 scale maps.

               iii. Cartosat-1 satellite is more stable compared to the other two because it is
               not continuously changing the view while imaging. This ensures that no
               scale variation will be there at different parts of the imagery and will give
               uniform accuracy over a single image as well as images taken from different
               orbits.

               iv. By exploiting the full potential of the payload and imaging geometry,
               high resolution data can be effectively used for large scale mapping with
               minimum control requirements.

               v. Better resolution satellites not necessarily give geometric accuracy
               required for large scale mapping. It is recommended to use stable satellites
               than agile for mapping applications.



                                             31
6.5.2   Discussion
        i. Maj Gen V S Deo and others raised the question of suitable DEM & Ortho Photo
    Grid size for scale specific products.
        ii. Sh G C Nayak Observed that image enhancement techniques other than
    radiometric adjustment (only modifying LUT), where original pixel values get
    modified, before Auto DEM generation which use the image correlation methods is
    fundamentally erroneous.
        iii. Sh Hanumanth Prasad opined that ortho photo may be generated in combining
    the LIDAR data.
        iv. The chair Maj Gen P N Koul said that ortho photo should be generated keeping
    in mind its requirements and demands of such products.

6.6 Working Group IV : Geospatial Data Structure
    Chair :           Maj Gen P K Vachher
    Presentation Lead by :    Maj Gen P N Koul
6.6.1 Presentation
       Maj Gen P N Koul began his presentation with a concern on the status of Digital
Photogrammetry in the department and looked beyond mapping in the geospatial data
context in digital environment where the demarcation between Photogrammetrists /
cartographers / scientists is rather blurred. Before dealing with the Data Standard /
Structure, he touched upon the basic data model (Vector & Raster) and different series
maps.

i. Need for Standard / Structure
       a. Standards have direct relevance to database standardisation and the data structure
       b. Improve efficiency/effectiveness/reduce costs
       c. Protect investment in data against technological changes
       d. Increased availability of more accurate, complete and the correct data

        He cautioned that different skills are needed to develop standards at each level and
people should work at a level in which they are most knowledgeable and have the requisite
aptitude.
        He remarked that Standards often require a long time to develop but good
standards are usually long lived.

ii. Scope of the Data standards / Structure
        It mainly encompasses
        a. Data Contents- may be primary data or derived data elements from primary data
like slope or composite map generated by overlap of various input layers.
        b. Unique Feature Codification for every input element.

iii. Salient Spatial Database standards
         The proposed standards, amongst various other aspects should address:
         a. Projection System
         b. Scale factor for various levels of database
         c. Co-ordinate system


                                            32
       d. Accuracy specification w.r.t various scales for both raster and vector data
       e. Minimum spatial Unit
       f. Meta Data standards
       g. Data dictionaries
       h. Lineage
       He presented a table of different scale data for possible purposes in different
hierarchy from village to national level.

iv. Spatial Database standards / structure

       Spatial Data has to be organized /structured in tabular form containing the
followings:
       a. Description and format of First Few Records
          (Record No., Field, Content, Type of format, starting byte, ending byte, remarks)
       b. Volume Directory Records
       c. General information Records
       d. Data Category Records
       e. Line Records
       f. Area Records
       g. Text Records

v. Present status of NSDB Standards in India

            (a)   DVD (DVD-1) of Survey of India in vogue for the last about 1 ½
                  decade

            (b)   NRIS standards of DOS

            (c)   NSDB Standards for Cartography- Committee on Cartography &
                  Mapping constituted by Secretary DST, under the umbrella of NNRMS
                  and represented by DOS and SOI have submitted the report in 2000.

            (d)   NSDE data standard & exchange format – NSDI adopted the NSDB
                  Cartographic Standards fully with scanty modifications which was
                  termed as NSDE, Data Standards.

            (e)   NRDMS standards – Following Sub committee for different categories
                  for 1:10,000 scale database were constituted.

       Boundaries                    KSCST
       Land use/Land Cover           JNTU
       Hypsography                   SOI
       Hydrography                   SOI STI/TN
       Transportation                KSCST
       Utilities                     Univ. of Hyderabad
       Building                      APSRAC & KSCST



                                             33
        Surveyor General of India have made Board of officers of Survey of India to
formulate Data Standards for 1:10,000. scale mapping with a aim to address the computer
friendly GIS ready data base where visualization aspects would be addressed separately.

vi. Loud Thinking

       He emphasized the concept of scale free data and suggested that both spatial and
non spatial be collected at 1:2,000 scale and deactivated / filtered the data for 1:10,000 &
1:25,000 scale with pointers.

vii. Salient stages

       The following stages are suggested :-
            (a) Arrive at a particular but Definite Data Structure.
            (b) Generate GIS ready databases.
            (c) Address “Visualization” separately by adopting the concept of style
                sheeting by use of Visualization parameters (VPs).
            (d) Develop interface/automation to derive the following possible “End
                Products” from the neat GIS data base. Parameters related to the
                Cartographic attributes & symbology, developed for each end product
                should reside in the interface.

       End Products:
             Topo Maps
             Utility Maps
             Project Maps
             GIS derived Plots etc

viii. Data Structure

       The following structure is suggested :-

   a. Major Categories

   I Settlements & Cultural Details

         1.    Hydrography
         2.    Ocean & Coast lines
         3.    Transportation
         4.    Land Cover & Land Use
         5.    Utilities
         6.    Government/Administrative Boundaries
         7.    Land Surface Elevation: Topographic / Hypsography
         8.    Geodetic
         9.    Vital Installations



                                            34
b. Sub Categories

1. Settlements & Cultural Details:

 •     Residential
 •     Commercial
 •     Government Offices
 •     Religious
 •     Antiquities
 •     Utility Centre
 •     Educational
 •     Institutions/Welfare/Relief

2. Hydrograph

 •     Stream
 •     Stream features
 •     River
 •     River Features
 •     Lake
 •     Lake Features
 •     Canal
 •     Canal features
 •     Other Water Features

3. Ocean & Coast lines

  •    Coast Line
  •    Coastal natural features
  •    Coastal artificial features
  •    Tidal
  •    Elevation data
  •
4. Transportation

 •     Carriageway
 •     Carriageway Infrastructure
 •     Tracks
 •     Transport Utilities
 •     Railway
 •     Railway Infrastructure
 •     Railway Embankment & Cuttings
 •     Airport




                                     35
5. Land Cover & Land Use

  •     Built up
  •     Agricultural
  •     Forest
  •     West Land
  •     Water bodies
  •     Wet Lands
  •     Grass Land/Grazing Land
  •     Snow Cover

6. Utilities

  •     Transmission Lines
  •     Pipe Lines
  •     Water Utilities
  •     Conveyor Belt
  •     Rope Way
  •     Filling Stations
  •     Solid Waste Management

7. Government/Administrative Boundaries

  •     International Boundaries
  •     State Boundaries
  •     District Boundaries
  •     Subdivision / Tehsil / Taluk / Mandal Boundaries
  •     Pragana Boundaries in UP
  •     Village Boundaries
  •     Cantonment Boundaries
  •     Municipal/ Corporation Boundaries
  •     Ward Boundaries
  •     Panchayat Boundaries
  •     Block Boundaries
  •     Constituency Boundaries
  •     Police Station Boundaries
  •     Park/Lawn Boundaries
  •     Boundary Infrastructure

8. Land Surface Elevation: Topographic / Hypsography

  •     Contours
  •     Mountain Features
  •     Mud Volcanoes
  •     Sand Features
  •     High Mountain Features



                                    36
       9. Geodetic

         •     Primary
         •     Secondary
         •     Topographic
         •     Other Sources

       10. Vital Installations

         •     Civil Vital Installations
         •     Military Vital Installations

       Detail data Structure is enclosed as Annexure II

ix. Conclusion

       He concluded emphasising the need of substantial thrust in adoption of digital
photogrammetry and formulation of data standards / data structures by harmonizing
various data structures in vogue.

x. View of Sh R Nanda

       Sh R N Nanda said that data structure should be flexible in such a way that it
should facilitate to come up with new version of structure to keep pace with the changing
scenario with the time and technology.

xi. View by Sh P V Rajashekhar

       Sh P V Rajashekhar said that data base should be GIS based rather CAD based and
need to rethink on visualization aspects. He also said that there should be one data model.

6.6.3 Discussion

        (i)    Maj Gen Tavinder Paul raised the issue of long pending decision on the data
               structure despite constitution of various committee.

        (ii)   In response Maj Gen P N Koul briefed the background of various
               committees and presented the status report.

        (iii) Sh Dhiraj Goel said that views of different organizations should be taken
              into consideration while preparing the Data model.




                                              37
6.7 Working Group V : 2D & 3D data capturing

    Chair :           Maj Gen V S Deo
    Presentation Lead by :  Sh S K Sinha

6.7.1   Presentation

        Sh S K Sinha presented the paper touching upon all the aspects of data capturing in
Digital Photogrammetry environment. His presentation was derived from his experiences
in Delhi 3D mapping project.

i. Choice between 3D and 2D data capture

a. Data can be captured in 2D mode from Ortho photos generated using freshly created or
existing DEM. The disadvantagees lies in using the existing DEM as these are;

    •   Burdened with errors in Z, due to different sets of Control Points used for Aerial
        Triangulation
    •   Burdened with limitations in scale of photography on which DTM based
    •   Burdened with errors in geo-referencing (X,Y) of DTM due to Datum Trans-
        formation and limitations in scale of photography.

        b. He justified the the data capturing in 3D mode citing the following advantages;
    •   Data with X,Y,Z has much greater value
    •   If accurate DTM is also a requirement, additional cost only for skilled manpower
        for 3-D Feature Extraction, as cost of input stereo imagery, Ground Control, Block
        Adjustment is same
    •   Additional effort for 3-D Feature Extraction is compensated by the savings in time
        and effort in digitization of breaklines and DTM editing
    •   2-D Digitization with ortho-photos is burdened with errors of
            – relief displacement in overground structures
            – Relief displacement in Building Foot-Prints
    •   Boundary walls and fences, delimiting plot boundaries in high urban areas are
        difficult to interpret in ortho-photos
    •   Features like electric poles, light poles, traffic signals, high mask light are difficult
        to spot in ortho-photos

c. he illustrated the effect of relief displacements in ortho photos and how view shed
analysis can be carried if data is captured in 3D mode.

d. He recommended the followings for the choice of options;
    • All original survey, specially on large scale should be carried out by 3-D
       Feature Extraction.        2D Feature Extraction should be used only for
       digitization of existing frame/non-frame maps/records .
    • 2D digitization may be used for subsequent updations



                                               38
   •   2D digitization may be resorted to off-the-shelf products to cater to specific
       requirements but should not be adopted to generate base topographical layers

v. Capturing methodology

       The flow diagram of the methodology 3D feature extraction is shown below :-




vi. Modeling of 3d buildings

a. The following Stages were illustrated :-

   •   Vector Capture as per procedure
   •   Validation – Topological Checks
   •   Establishment of Internal Topology and building of Polygons
   •   Texturing with geo-located and geo-referenced images, model wise
   •   Integration of Terrain Data, Vertical Imagery and 3D Textured Models




                                              39
   b. The following aspects of 3D building are to be captured :-

   •   Building outline
   •   Courtyard
   •   Building break lines
   •   Address divider
   •   Roof spot
   •   Roof ridge
   •   Roof point
   •   Slope roof line
   •   Curved roof line
   •   Obscured area outline
   •   Onion dome
   •   Top of onion
   •   Top of tube
   •   Awning

c. Limitations of 3-D modelling of buildings are :-

   •   can't create line work to model negative space (Onion Dome is exception)
   •   not intended for engineering purposes-- level of generalization
   •   not intended for wire frame creation
   •   shadows/trees/building lean/tall buildings prevent 100% capture
   •   must have 80/60 overlap to see all building detail

d. Guidelines for 3D Modeling of Buildings :-

       Building

   •   be precise with corners and angles
   •   what should be captured as a building:
   •   large commercial tanks, chimneys

       Building Roof Spot

   •   capture roof spot in center of polygon
   •   if other than flat roof, elevation of spot should be average elevation of polygon
   •   highest polygon gets roof spot using top of highest wall
   •   all lower polygons should place roof spot on actual roof surface

       Address Divider Line

   •   obvious changes in address (based on roof architecture/color/texture)
   •   must be snapped to the Building Outline
   •   takes precedence over the Building Breakline
   •   address of elevated walkways that connect buildings belongs to only one building



                                             40
       Building Breaklines

   •   linework "on" feature as much as possible (helps qc & quality)
   •   top-down compilation... easier for the eye to see and measure
   •   keysnap as much of the data as possible... most accurate
   •   place vertex at location of future snapping to speed compilation
   •   min area for polygons (2m x 4m)... unless important feature
   •   3 meter minimum change in height between polygons
   •   walls not captured as polygons unless 1m in thickness

e. Followings are the two tips for better result

   •   set other stereo pairs for better visibility of details
   •   review line work in 2-d to check that all polygon contain 1 label

vii. Optimum zoom level

       The optimum zoom level in stereo mode is 1X – 2X and the threshold level in 2D
mode is when pixel breaks.

viii. Contour generation

a. Contour can be generated in following three ways :-

   •   Manual Method where in manual contours can be captured in 3D from stereo
       models.
   •   Semi Automatic Method by extracting break lines and mass points from the terrain.
   •   Automatic Terrain Extraction (ATE).

   However in semi automatic method with break lines complex terrain can be handled to
obtain a better result.

b. Following features should be introduced as break lines for contour generation :-

   •   Features to be extracted for DTM are:
   •   Road (line):- Only Paved and Continuous road or width more than X mts to be
       captured.
   •    River (line):- Capture the entire stream, river (single and double line (5mts width),
       drain, ditch, and canal.
   •    Break Lines: - Break lines to be captured in all the cutting, embankment, retaining
       walls and sudden change with X mts elevation difference in terrain.
   •   Lake, lagoon (shape):- All the water body needs to be captured in present water
       area with fixed height.
   •   Urban area (shape):- Capture the settlement area in ground level without crossing
       hedge, fence or any permanent feature. Place spot height in the street crossing



                                            41
       inside urban area. Breaklines around perimeter of all buildings that straddle model
       neat line
   •   Bridge or viaduct (shape):- Capture all the permanent bridge with road elevation in
       shape. Place a spot height (Spot Height on Bridge) on the bridge. breaklines on
       road should be squared off at ends of bridge (helps ortho)
   •   Rail(line):- Capture the entire railway with double linebreaklines @ big
       tops/bottoms
   •   Top and Bottom of hill tops, ridge lines

c. Following Spot height should be considered :-
    • Local high points (summits)
    • Local low points
    • Saddles
    • Lowest points of quarry pits
    • Lowest points of excavations
    • Along valley beds
    • Standing water bodies
    • Street intersections
    • Along the centerline of roads
    • Center of every bridge and such other appropriate locations.

ix. Quality issues

a. Quality Assurance and Quality Control

       The following aspects should be addressed :

   •   Documentation of Process, QA/QC Procedures
   •   Listing of (Quality Control Records) QCRs at each stage
   •   Multi-stage Quality Assurance
   •   Responsibility and Accountability for Quality Audit at each stage to be clearly
       assigned
   •   QCRs to be generated at each stage to be properly archived with date and stamp
   •   Normal Audit should be followed by Tightened Trail in case of doubt/problem
       areas

b. Method of scrutiny

   •   Check individual and associated layers Layerwise for omissions and consistency
       with underlying rater/models
   •   Topological Validation
   •   Edge Adjustment




                                           42
c. Error evaluation

   •   Sharp well distributed check points whose x,y,z are known and identifiable on
       imagery / model should be measured.
   •   Stamping the statistical accuracy with DTDB, maps etc.

x. Out Turn

        From the experiences in the DSSDI project he stated he stated the achievable out
turn for a specific task as shown below;

       Scale of Mapping: 1:10K
       DTM grid interval: 10m
       Accuracy: 50 cm

      3D Feature Extraction (Excluding Buildings): 0.2 sq. km / man day
      3D Buildings for modeling: 0.2 sq km / man day
      DTM: 1 sq km / man day
      Ortho: 5 sq km / man day

xi. Record maintenance

a. Quality Control Records

       Quality Control Records will refer to

   •   system generated reports indicating the results achieved in processes,
   •   inspection notes by Quality Auditors
   •   Stage approval by Quality Auditors of the Employer or the Contractor.
   •   QCRs can be both in soft and hard copies
   •   Preferably in pre-designed proforma, tables.

b. QCR for DTDB

   •   Progress Report
   •   Inspection Report on Practices and Processes every 15 days – on the checklist
       devised.
   •   Validation Report on Topology.
   •   Inspection report for s/copy examination for completeness, matching with
       underlying raster / model, symbology
   •   Edge Adjustment Report
   •   Inspection report of H/copy examination




                                               43
xii. View of Representative from IIRS

       Ms. Shefali Agrawal said that the choice of 2D or 3D data capturing depends on
the type of terrain. In a plain area data capture in 2D environment may serve the required
purpose but in a complex terrain with lot of height differences, data must be captured in 3D
environment.

6.7.2   Discussion

       As the presentation was crisp, to the point and based on project experiences there
were no questions of any substance from any quarters.

        The chair concluded the session thanking Sh S K Sinha with a note of appreciation
for his very good presentation.

6.8 Comments from Industries

       The session was chaired by Maj Gen (Dr.) B Nagarajan. He invited the
representatives from other organization and industries to present their opinion on the
subject and requested all participants to be open minded so that the development going
through the technology all around may be appreciated for its effective implementation.

6.8.1   Infotech Enterprises Ltd.

        Sh Hanumanth Prasad introduced the latest development in the technology and the
level of automation and improvisation being made in the production line in his presentation
under title „Changing Trends in Digital Photogrammetric Techniques - A Production
Perspective’. In his presentation talked about the Bingo / Aerosys S/W being used in
Infotech Enterprises Ltd.

a. Transformation

        The following Transformation is taking place in Digital Photogrammetry;

    •   Aerial Triangulation with manual point selection – use of automatic point selection
        and autocorrelation combined with Bingo/Aerosys bundle adjustments for precise
        accuracies.

    •   Basic photogrammetry is to use aerial films - The present trend is to use digital
        images such as DMC, ADS40, Ultra Cam, or video sensors.

    •   Interactive collection of mass points with breaklines for creation of Digital Terrain
        Models – The present trend is to use LiDAR data for generation of contours and
        orthos and advanced Automatic DEM generation process with latest software.




                                             44
   •    Repeated collection of survey ground control points and digital terrain model data
        for digital orthophotos – The present trend is to re-use all the existing data in the
        process called “Second Generation Digital Orthophoto”.

   •    Low resolution Landsat and other satellite imagery - The current trend is to obtain
        high resolution (60 cm) high accuracy satellite imagery and RADAR Sat data.

b. Digital Aerial Triangulation

        Since 2001 the level of automation and volume of data handling during Aerial
Triangulation has increased many fold. We have standardised the bundle adjustment
process for image count less than 1000 in Aerosys and between 1000 to 10,000 under
Bingo software. These processes are time bound and results are achieved with accuracy in
a short span of time providing additional parameters for Quality Control and analysis.

c. Digital Data Mapping

        Data mapping is the second important task where technology and expertise are
required to meet the standards and the quick turnaround time expected by the industry.
Following are the key factors that would enhance the ability of any Mapping service to
create a value addition in the industry:
      • Right Selection of data as per the user requirement
      • Technology interface on the available software platforms / Tools development
      • Efficient Project management
      • Defined Process and procedures
      • Continual Improvement
      • Innovative ways of approach on the process
      • Providing solutions than posing problems
        The Project Management Information System (PMIS) develped has been embedded
into the process of photogrammetry. The integration of PMIS along with tool development
process has emerged into a new trend of production process which enhanced both speed
and quality.

        Infotech uses sophisticated tools and wide variety of techniques to execute fast and
high-quality planimetric mapping. Utilizing a variety of platforms within the GIS and the
CAD environment, Infotech has developed tools that enhance not only the speed of data
collection, but also check the output for a consistent quality data.
Following are some of the tools that are highlighted

       Planimetric Tools

               Linear and Polygon features digitization:
               Streams digitization:
               DTM Grid
               Spot Label
               Building Tool – Auto Orthogonality and Ortho closing



                                             45
     MapEdit Tools

               Break / split Elements
               Cell – Rotate
               Cell – Scale
               Cell - Replace
               Deleting Masspoints vs Breaklines
               Quality Check Tools

d. Ortho Photo Generation

        With the advantage of high quality and high precision of digital images, the
production speed of ortho photos will be high and this in turn leads to a most cost-effective
method for photogrammetry production environment.

       The orthophoto generation process can be carried out separately for the images
having tall buildings with different angles. The ortho photo is generated using draw
polygon method i.e. selecting each independent cluster of buildings with different leans.
All the generated orthos with the same lean and shadow properties, will be mosaiced in
Orthovista.

        The main characteristic property in this type of ortho generation is compensation of
occlusion by using patch of ortho image from adjacent ortho pair to overcome the lean and
as per the cutline.

       Key areas for Automation - Enhances the Ortho Handling:
          • Automatic DTM Generation
          • Ortho photo generation in a Batch mode with sequential
          • Exclusion of shadow areas using 3D models
          • Auto Seam
          • On fly corrections for radial distortions
          • Photoshop routines for radiometric balance.

e. Digital Elevation Model

       With the latest trends in using LiDAR data, the processing time has been drastically
reduced. We have been integrating successfully LiDAR in various large-scale mapping
programs across the globe for generation of Orthos and Contours.

        Using the LiDAR technology, one can acquire thousands of square kilometers in a
less than a day‟s time. Hence, the collection of mass points for creation of new DEM is no
more required and break line collection is limited.

       Using LIDAR data we can classify for Bare earth DEM and city surface models.
       Some products generated from LiDAR data were shown.



                                             46
f. Typical Photogrammetry Production flow




g. Satellite Ortho Rectification

       • The sophisticated geometric and radiometric characteristics of high spatial
       resolution satellites like Cartosat and IKONOS sensors provides the end user with
       excellent metric accuracy and wealth of information
       • Infotech has made practical attempts to create 5mt DEM from Cartosat imagery
       to Ortho rectify 1mt IKONOS data for further process of data mapping and GIS
       creation.
       • The blend between expertise and software technology has made this possible to
       achieve these high precision product as uniform seamless mosaic.
       • well-defined and distributed control points accompanied by Rational
       Polynomial Coefficients enable us to get accuracies with no discernable loss
       • Software such as SOCET SET, ERDAS, and PCI Geomatica can be used to get
       these high-precision products with uniform and seamless mosaics. These high-
       precision orthos are used for large-scale mapping with high accuracy.

h. Recommendations for Training Curriculum

       •   Digital photogrammetric techniques for AT and Ortho generation
       •   Stereo satellite ortho rectification
       •   Advanced methods of DEM / DSM generation – A special emphasis on LiDAR
           / RADAR or IFSAR


                                          47
        •   Real definition of a Mapping process
        •   Data Formats and Meta data generation ( Supported Mr. SID format)

Conclusion

       With the advent of advanced technologies and high quality imagery and
implementation of right process approach coupled with good Project Management
Information Systems, any Mapping service organization can serve the customers around
the world with highly accurate products delivered on time.

        In the present world of constant changes, engineering service providers with
exposure to a wide variety of technical environments, would be able to successfully and
efficiently cater to various segments of the industry.

6.8.2   National Remote Sensing Centre

       Dr. Venugopal Rao claimed the successful use of stereo satellite imagery in
preparation and updation of Geospatial data base on 1:10,000 scale for various project
works. He also said that DEM from Cartosat 1 can be used for orthorectification of
Cartosat 2 and orthorectified image can be used for updation of large scale map prepared
from aerial photographs.

      According to him seamless geospatial data base for 3,000 sq km was completed in
3 months.

        Topographic maps of Indian coast was from Cartosat 1 data was prepared using
Digital photogrammetry. In the process DEM was generated with 10 m posting. Relative
accuracy of 2m and absolute accuracy of 5-6 m was achieved. Till now 1.3 lakh sq. km.
area is completed.

       His strong suggestion was that standard and requirement may be separated for
quick delivery of the data to meet the demands of purpose like Disaster management and
Rural Development.

       Though some questions were raised on the compromise of the standard the chair
intervened with a request to consider various views on providing the solution.

6.8.3   Speck System Ltd.

        Sh J N A Murthy presented on the growing demands of the data pressure on the
industries from various clients. He emphasized that delivery of products in time is the key
factors to sustain in the business. To meet the time line standard has to be compromised as
per the requirement of the project like Municipal GIS only for taxation purpose, Urban
Planning need just plan, Corridor mapping for infrastructure etc.




                                            48
         He said that the volume of jobs are enormous and cited some of the projects like
irrigation projects and revenue maps of 600 districts. To carry out such a task industries
need the geospatial technician. In this regards he look forward towards Survey of India to
build the capacity in this field. He opined that Joint project between SOI and NRSC will
help in providing the solution.

6.9 Panel Discussion

        Following nominated members were introduced by Maj Gen (Dr.) B Nagarajan and
invited them to the dais to initiate the Panel discussion on ‘Effective implementation of
Digital Photogrammetry in SOI’. He put before the panel the issues of choice of Input
Data, use of products like LiDAR / IFSAR and Geoid model / Vertical Datum.

               1. Sh. G S Kumar, Chairman
               2. Maj Gen Manoj Tayal
               3. Maj Gen Tavinder Paul
               4. maj Gen S Ravi
               5. Sh P V Rajashekhar

        The chair of the panel thanked for giving the opportunity and explained the
importance of the topics in the present context. He invited the members one by one to give
their views.

6.9.1   Sh P V Rajashekhar

        He stated that SOI is confused in setting its priority to carry out its task effectively.
It has the capacity to implement any technology, only the question is to recognize its
strengths. He cited the strength of provision of ground control and survey. Given the
opportunity and liberty the staffs are capable of raising to high standard. In the past the
technology intensive tasks has been completed successfully by the existing, not so
qualified, staffs.

       In the case of digital data base cartographic visualization can be handled with the
GIS ready data by adopting and customizing the line style so that repetation of sybolisation
process can be avoided.

6.9.2 Maj Gen Manoj Tayal

       The Workshop has given an opportunity to heard many new idea in adopting the
technology. He hoped that the WG should give the report for consideration in SOI. He
expressed the concern of getting the technology superseded very quickly over a time
period which makes the organization difficult to switch over from one to another.

       In respect of Digital Photogrammetry he said that it must be implemented to its full
potential in SOI and it must be done first at IIS&M. Any new technology need to be
looked into and develop it in collaboration other organization.



                                               49
      He expressed the need of defining a precise Vertical datum / Geoid model citing the
example of problem faced during the Mumbai project.

        There is a requirement to implement a project in time bound manner. However the
strength of SOI lies in its inherent standard, which should not be compromised.

6.9.3 Maj Gen Tavinder Paul

       He began with aopening remark that being in the service for more than 30 years in
SOI his views are bound to be SOI centric. He said that it was too late to organize this
Workshop when the technology has been used so extensively outside.

        He tries to find the reason for not implementing the technology in SOI effectively.
In this he stated that there is a structure in all GDC level to carry out the basic process of
mapmaking which was implemented in analogue environment. However it started
crippling in 1990s because of financial crunch which effected the technical strength also.
So naturally the chain in the process of map making, which involves the photogrammetry,
broke. Thus it effected the implementation of the new technology in digital data
acquisition in the absence of new job. However collection of basic data from field survey
can not be dispensed with, which is the strength of SOI.

       He expressed his concerned of shortage of man power to carry out the mapping in
1:10,000 scale when task of completing 1:25,000 scale mapping is still laying.

        He said that the data of indigenous satellite must be looked in to as an input in
creating large scale data base on project basis for timely delivery of the products.

6.9.4 Maj Gen S Ravi

        He supported the use of satellite imagery wherever possible and suggested for joint
projects with Space Organisations.
        Regarding the use of LiDAR data he commented that it has been tested only in
Delhi area once in India.

        He was categorical to say that as a national Mapping agency SOI can not dilute the
established standard.

        There should be solution to derive the precise height using GPS observation to
avoid leveling work.

        He commended the tremendous contribution of SOI staff in carrying out the
assigned tasks despite minimum training. He stated that though the volume of work is
increasing staff size of SOI has reduced a lot over a time period.

       DEM from the existing data will be most useful for various purpose.



                                             50
6.9.5 Comment of Chair

       The chair briefed the comments of various members and put forth his own view
before opening the session for discussion.

       There is a great demand of the base data which need to be provided in time to fulfill
the geospatial data. In the creation of data base for various application there are a lot of
duplication work being carried out by various organization due to lack of coordination and
cooperation. His view is that this must be avoided through proper coordination.

       We should provide the product what technology supports and should be careful in
using hybrid technology.

6.9.6 Discussion

       (i)    Sh N K Agrawal stated that State Survey Department should be equipped
              enough to take up the task.

       (ii)   Sh P K Das said that IIS&M is contributing towards the training in the field
              of Digital Photogrammetry and Personnel trained in this institute are
              performing well in various industries.

       (iii) Maj Gen KRMK Babaji Rao recommended to consider following two classes
             of mapping;

               (a)   Project Mode Mapping – to be completed within a time frame as per
                     the user requirement & available budget.

               (b)   National Mapping – Based on the uniform standard, accuracy and
                     methodology.

       (iv) Sh T C Neelakanthan appreciated the conduct of such Workshop which was
            an education for him.

              He strongly objected to the comment that SOI is confused, rather in his view
              the decision and policy makers in higher up in the government are confused a
              lot to provide clear guidance to the department.

              He commented that SOI need not feel despondent seeing the fast adoption of
              technology and quick delivery of products by industries. SOI has a definite
              national role to play, it should not reduce itself to play the role of contractor.

              IIS&M has a certain role to play in taking forward the technology.




                                              51
(v)   Maj Gen P N Koul said that SOI has switched over to various technologies at
      different stages successfully in the past. There is a system in place at GDCs
      level. Training can be organized at IIS&M and GDC levels to implement the
      technology successfully.

      The surveying tasks can be taken up in two modes,

               (a) Project mode
               (b) Departmental mode.

(v)   Dr. Malleshwar Rao raised the issues of storing and handling of huge data
      involved in Digital photos / images. Security aspects has to be addressed
      while sharing the acquired data. For that the required infrastructure need to be
      built across the organization.

      He also raised the issue of integrating the existing data and how to publish
      and disseminate it for public use, whom it is meant for.

(vi) Sh Hanumanth Prasad recommended the following :-

       (a)   3D data capturing for initial survey and 2D capturing for updation

       (b)   !:5,000 – 1:500 scale mapping from Aerial photography and 1:10,000
             and smaller scale mapping from Stereo satellite imageries LiDAR
             data . In specific cases Ortho rectified mono satellite imageries can be
             used.

(viii) Dr. P V Radhadevi said that there is a need of precise Geodal undulation to
      rectify the satellite ephemeris which will help in adopting hybrid approach.

      Updated 1:50,000 and 1:25,000 SOI topographic maps will help immensely
      to register an image for various applications. She recognized the role of SOI
      in providing the GCPs.

(ix) Dr. Venugopal Rao suggested for cooperation amongst various organization
     to avoid any duplication of work and coming together for formulating
     methodologies.

(x)   Dr. S K Singh emphasised the need of defining the Geodal model which help
      in converting the height observed through GPS ( Ellipsoidal Height) to
      Geodal height which is obtained by leveling. This will reduce the work of
      leveling which is a time consuming and costly. This exercise has already been
      tested for 4º X 4º area in different part of the country found the results to be
      consistent.




                                      52
       (xi)    He suggested for defining the standard for different grade of products. SOI
              has to look for long term plan.

              Large scale mapping is doable if manpower and money is available.

               Spatial data is to be treated as Infrastructure and SDI approach is there in
              place for data sharing.

       (xii) Ms. Shefali Agrawal expressed that Technologies are complimentary and
             should not be taken in apprehension.

              For effective implementation man power development is essential.

       (xiii) Sh R N Nanda said that in 1997 a survey on demands of digital geospatial
              data was conducted and found that there is demands of 1:4,000 and 1:8,000
              scale data.

              The scale of data acquisition should be decided first and then the
              methodology will follow. And then accordingly departmental training can be
              focused to develop the man power for specific task.

       (xiv) Sh B K Mahapatra said that all GDCs are having the infrastructure and trained
             man power. They are capable of carrying out the jobs. Due to other priorities
             and jpb pressure on a limited man power the digital photogrammetry could
             not be implemented effectively. There is need to define the standard for every
             process and specific SOP for effective implementation.

       (xv) The chair expressed his view that we should not be constraint by standard and
            have to compromise some where as per the technology and requirement of the
            purpose and users.

7. Concluding address by Additional SG IIS&M

       Maj Gen (Dr.) B Nagarajan clarified between GPS and DGPS, generally used
loosely used in tandem. He requested to understand the differences and don‟t go by the
vendors version on this.

       He explained the efforts of IIS&M in building the capacity in the field of geospatial
technology and it is in line to make the institution an university so that recognized degree
and post graduation certificate can be awarded.

        Lastly he thanked one and all, the participating organizations, industries,
participants, organizers and volunteers who were involved in the successful conduct of the
Workshop.




                                            53
8. Summary

8.1 Choice of Input Data
       i. For 1:10,000 and larger scale maps / utility maps with Contour / precise height
information and Revenue maps aerial photography following the established norms should
be used.

       ii. Image obtained from aerial digital camera, when available, must be evaluated for
on project mode to find out the scale of output product that can be generated.

       iii. Utility of LiDAR data for height information must be evaluated in different
condition before integrating with Stereo Aerial / Satellite image data.

         iv. For 1:15,000 and smaller scale maps with contour interval 10 m or more
satellite stereo image of Cartosat 1 , which has better orbital stability in comparison to
other similar satellite, should be used. However in case basis when only plan data are
required it may be tried for 1:10,000 scale maps also.

        v. While choosing the stereo satellite imagery products like World view, Geo Eye,
etc., where supplied with multi images, the number of stereo pairs required, overlapping
and cost involved must be evaluated.

        vi. Ortho image generated from Cartosat Stereo satellite imagery and Ortho
rectified high resolution satellite imageries may be used for updation of existing data base
on 1:15,000 or smaller scale.

       vii. Multi Spectral mono image should be used for Thematic mapping.

8.2 Provision of GCPs
         i. For aerial photos of frame camera with out any IMU information the established
pattern of GCP distribution; maintaining the bridging distance in every five models apart
and in all lateral overlaps in the periphery of the block, must be followed. Two GCPs each
in the over lap areas in first and last models of first and last strips should be provided,
where GCPs in the lateral overlap area must appear on more than 4 images.
         The number of GCPs should not be less than 10% of the number of exposure
station.
         All GCP should be of Plan & Height control as far as possible.

       ii. All models adjoining gaps or large water bodies should have one GCP.

       iii. Aerial photographs with IMU information and images from aerial digital
camera in different terrain should be evaluated for pattern and optimum number of GCPs
requirement in project mode.




                                            54
       iv. In case of stereo satellite image with RPC data 5 GCPs ( 4 in corners & one in
middle) should be provided if area is covered by single pair. If area is covered by multi
image pairs 9 GCPs in 3 bands should be provided.
       v. The number of check points should not be less than 10% of the total number of
GCPs and should be evenly distributed through out the block providing in the strip overlap
area. One must be at the middle of the block.

        vi. GCPs must be post pointed in a print of enlarged view of the image being used
in the processing for unambiguous identification in the lab. It is suggested to support with
a digital photographs of the point taken at the site.

        vii. During project planning stage some GCPs may be pre pointed depending on the
project specification putting appropriate signals.

       viii. GCPs should be provided through GPS observation. Maximum leg distances
from the base station, established by 72 hours observation, should be within 20 km and
observed for 1 hour. For height determination these should be connected with the BMs
with maximum distance of 5 km.

       vii. Accuracy standard of GCPs should be better than ½ pixel of the input image in
terms ground distance.

 8.3 Aerial Triangulation (AT)
       i. Distribution of tie points should be min 3 columns X 3 rows in each photo.. For
tough areas 5 columns X 5 rows in each photo is recommended.

        ii. Central tie point to be on min of three images. Upper and lower points to lie
within the lateral overlaps of the strip. Each tie point to be on six photos as far as possible.

       iii. Often scanned input data are of poor quality where tie points generated
automatically need scrutiny which is time consuming, tedious and required expertise. In
such cases it is recommended to measure manual tie points. It is suggested to try with a
combination of manual and automatic approach and may be generated in phase wise (Strip
or Model wise) for a large block with complex input data.

       iv. Parameters for the processing of AT may be set as follows:
               Standard Deviation of the Image measurements: 1/3 of the pixel
               Standard Deviation of the Controls Points: Fixed / as per the GPS survey
               No of Iterations: 10
               Convergence : 0.00010 meters / project specific unit
               Blunder detection can be time saving/robust blunder check can be used for
               identifying blunder points.
       RMSE for GCPs & Check points(in Ground & Image coordinates term)
               Total Image unit – weight RMSE (Sigma 0) < 1 Pixel
               Root Mean Square Error (RMS) of residuals in X,Y & Z at the Full controls
               < 1 Pixel



                                              55
               Root Mean Square Error (RMS) of residuals in X,Y & Z at the check Points
               < 1 Pixel
               RMS value of residuals in image coordinates < 1 pixel

8.4 Generation of DEM and Ortho image
        i. Orthophotos are not substitute for line maps. These can be used with caution as a
substitute for a planimetric line map for certain smaller scale, nondesign applications.
        ii. Time, cost , purpose of DEM ( need bare terrain or surface data) and required
accuracy level are the deciding factors for the procedure to be adopted ( Automatic or with
manual mass point and break lines), sample density and grid size.
        iii. The enlargement ratio from the original photo to output ortho photo may be
from 1:4 to 1:10 which is dependent on requirement, types of terrain and accuracy level of
DEM.

8.5 Data Structure
       i. The generic data structure organized for different scale mapping, enclosed as
Annwxure II is recommended for consideration.

         ii. It is suggested to develop the strategy / algorithm employing the professional to
filter the scale specific data base form the original base data captured as scale free.

       iii. It is suggested to address the symbolic visualization of GIS data base by
adopting line styling to avoid symbolization process.

       iv. It is suggested to develop interface / automation process to derive the possible
“End Products” like Topo Maps, Utility Maps, Project Maps, GIS derived Plots etc.from
the neat GIS data base. Parameters related to the Cartographic attributes & symbology,
developed for each end product should reside in the interface.

8.6 2D and 3D Data capture
       i. All original survey, specially on large scale should be carried out by 3-D Feature
Extraction to generate base topographical layers and 2D Feature Extraction should be used
only for digitization of existing frame/non-frame maps/records . 2D digitization may be
used for subsequent updations and off-the-shelf products to cater to specific requirements.

       ii. The optimum zoom level in stereo mode should be kept at 1X – 2X and the
threshold level in 2D mode is when pixel breaks.

      iii. In complex and built up area DTM should be generated by manual mass point
measurements and including break lines to generate the contours.

        iv. Quality of the captured data should be scrutinized layer wise, topologically
validated, checked for edge matching and errors will be evaluated by taking well
distributed check points.
        v. Quality Control Records (QCR) for all the process must be documented.




                                             56
9. Recommendation
        Based on the deliberation in the Workshop followings are recommended to the
department for consideration so that the Digital Photogrammetry is implemented
effectively to its full potential.

9.1. Rethinking on Products

9.1.1
        SOI must rethink on generating two class of products to meet its core mandate as
well as the requirements of various application of national importance.

        a. Topographical maps and other committed products of established standards and
           specifications generated following the tested methodologies.
        b. Requirement / Purpose / Project specific products to be delivered within a given
           time line using the available hybrid techniques where standard is not very
           stringent..
9.1.2
       SOI need not venture in to 1:10,000 scale topographical mapping for whole
country. Instead a large scale updated DTDB may be generated on 1:25,000 or 1:15,000
scale depending upon the area using existing 1:25,000 scale DTDB and satellite stereo /
mono imageries.
       Hybrid data like orthorectified satellite imageries generated using DEM from
various sources including LiDAR should be explored for generating such database.

9.2 Pilot Project
        Each GDC should be allotted a pilot project to generate large scale / medium scale
data base following the recommendations and suggestions in respect of input data,
provision of GCPs, Data capturing, and data structure. The projects should be monitored /
coordinated centrally and report so made by the GDCs should be circulated to all GDC for
their comments and suggestions for improvisation.

9.3 R & D cell

        An R&D cell may be located at IIS&M to review and update the SOPs for
generation of new products. All GDCs should coordinate with the R&D cell on the pilot
projects, furnish their feed back based on their project experiences and explore the
possibility of customizing the process as per the departmental requirements.

9.4 Capacity Building

9.4.1 Development of IIS&M
       Faculties at IIS&M should be exposed to the development in the technologies
through training at and collaborative projects with appropriate institutions / Organisations.
The training methodologies and curriculum may be redesigned to meet the present
requirements. For such collaboration SOI may enter into MOUs with institutions like
IIRS, ADRIN etc and selected industries.



                                             57
9.4.2 Development of GDCs
        It should be mandatory for all GDCs to train their manpower of Operator,
Supervisory, and Executive level in the technology. SOI should chalk out the time bound
program impart training for these levels and allot the jobs to engage the trained manpower
in Digital Photogrammetry.

9.5 Development of Facility
       A limited licences of Digital Photogrammetry S/W like Socketset, Bingo etc, which
provide fast solution and capable of processing large volume of data may be installed at
some selected zones.

9.5 Development of Infrastructure for data archival
       Infrastructure with required H/W and S/W may be developed along with security
and safety measures to archive the huge volume of scanned / procured images. A policy
guidelines need to be defined for cataloguing, maintaining, sharing and transporting the
image data.




                                           58

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:55
posted:12/18/2010
language:English
pages:58
Description: Evolution of Photogrammetry Cross Currency Pairs