Get documents about "very useful for historical studies and
Document Sample
Fundamentals of
Remote Sensing:
Digital Image
Analysis
Advanced Remote Sensing:
Introduction to Digital Image Analysis
Lecture 1
Prepared by R. Lathrop 10/99
updated 9/03
Readings:
ERDAS Field Guide 5th ed
ERDAS CH. 1: 1-15; 3: 52-77
Center for Remote Sensing and Spatial Analysis, Rutgers University
Lecture Notes 1: Overview of Remote Sensing
A number of these slides were originally
produced by Scott Madry and Chuck
Colvard with some subsequent
modification by Rick Lathrop. Additional
slides were produced by Rick Lathrop.
Center for Remote Sensing and Spatial Analysis, Rutgers University
Remote Sensing
• “Remote sensing is the science and art of
obtaining information about an object, area,
or phenomenon through the analysis of data
acquired by a device that is not in contact
with the object, area, or phenomenon under
investigation”.-Lillesand & Kiefer (1987)
Center for Remote Sensing and Spatial Analysis, Rutgers University
Radiation source/target/sensor
Electromagnetic energy interacts with
physical matter in different ways in
different parts of the spectrum.
Some energy is scattered, absorbed, etc.
Center for Remote Sensing and Spatial Analysis, Rutgers University
Digital Image Acquisition & Analysis
• Digitization of analog aerial photography, can be
very useful for historical studies and/or for high
spatial resolution needs
• Direct acquisition using some form of digital
imaging sensor
• Computerized image analysis can help to enhance
and extract information content of imagery in a
time-efficient, cost-effective manner
• Computers still can not replace the human image
analyst
Center for Remote Sensing and Spatial Analysis, Rutgers University
Aerial Cameras
A large format oblique Keystone’s Wild RC-
camera 10 mapping camera
Aerial photos
• Black & White - single
panchromatic layer
• Color: 3 layers B-G-R
• Color IR: 3 layers G-R-
NIR
Center for Remote Sensing and Spatial Analysis, Rutgers University
Aerial photos
The traditional form of remote
sensing
Pro:
• Can be easily customized to
meet specific requirements
Con:
• Can be expensive
• Need access to plane
• Time consuming interpretation
• Repeat coverage often
infrequent
• Different sun angles
Center for Remote Sensing and Spatial Analysis, Rutgers University
Space-borne Remote Sensing
Emerging Technology
Pro:
•GIS ready
•faster turn around
•acquisition time of 5 minutes gives
equal solar illumination, shadows,
no clouds
•easy to repeat for change detection
Con:
•Significant investment in computer
hardware/software
•Less flexibility in acquisition
Center for Remote Sensing and Spatial Analysis, Rutgers University
Passive electro-optical systems
• Electronic sensors can acquire data
outside the visible spectrum
• Elements sensitive to electro magnetic
energy (EME) of certain wavelengths
focus energy onto a sensor plane. A
prism is used to divide the energy into
specific wavelengths. The CCD’s are
stimulated and produce an electrical
signal equal to the energy focused upon
it. These data are recorded.
• Data are processed and displayed on
computers-images are composed of
“pixels”, whose brightness relates to the
strength of the radiation received from
an area on the surface.
• Digital processing of the data produces
useful information
Center for Remote Sensing and Spatial Analysis, Rutgers University
Design of A Remote Sensing Effort
• Clear definition of the problem
• Evaluation of the potential of remote sensing
techniques
• Identification of appropriate remote sensing data
acquisition procedures
• Determination of the data interpretation
procedures
• Identification of the criteria by which the quality of
information can be evaluated
Center for Remote Sensing and Spatial Analysis, Rutgers University
Resolution
• Four kinds of resolution determined by user needs:
• Spatial Resolution: How small an object do you need
to see (pixel size) and how large an area do you need to
cover (swath width)
• Spectral Resolution: What part of the spectrum do you
want to measure
• Radiometric Resolution: How finely do you need to
quantify the data
• Temporal Resolution: How often do you need to look
Center for Remote Sensing and Spatial Analysis, Rutgers University
Spatial resolution
Instantaneous Field of View (IFOV)
determines the dimension, D,
of the Ground Resolution Cell
(GRC) imaged on the ground
IFOV
Center for Remote Sensing and Spatial Analysis, Rutgers University
Spatial
resolution
keeps
getting
better...
Center for Remote Sensing and Spatial Analysis, Rutgers University
Spatial resolution
Center for Remote Sensing and Spatial Analysis, Rutgers University
1, 3, and 10 meters
Center for Remote Sensing and Spatial Analysis, Rutgers University
ultra-high spatial resolution
• 24 inch (60 cm)
• 6 inches (15 cm)
Center for Remote Sensing and Spatial Analysis, Rutgers University
Swath width
Landsat-185km (100 mi)
• 80 m = 40 Mb-4 bands
(MSS)
• 30 m = 320 Mb-6 bands
(TM)
• 10 m = 342.25 Mb-1band
• 5 m = 1.369 Gb -1 band
• 1 m = 34.225 Tb - 1 band
How small do we need?
How much data can we 185 by 185 km
store and process?
Center for Remote Sensing and Spatial Analysis, Rutgers University
Spectral Resolution: slicing up the
electromagnetic spectrum
Center for Remote Sensing and Spatial Analysis, Rutgers University
The electromagnetic spectrum
Comparative Sizes: from subatomic to human scales
Atom
Nucleus Molecule
Pinhead Human &
larger
Atom Bacteria Honeybee
From NY Times graphic 4/8/2003
Center for Remote Sensing and Spatial Analysis, Rutgers University
Spectral wavebands of Landsat TM
Center for Remote Sensing and Spatial Analysis, Rutgers University
Landsat TM-7 bands-8 bit data
Spectral Landsat TM BAND 1 2 3 4 5 7 6
(where we look)
Radiometric
(how finely can we
measure the return)
0-63, 0-255, 0-1023
Center for Remote Sensing and Spatial Analysis, Rutgers University
Landsat TM: each waveband provides different information
about earth surface features
Center for Remote Sensing and Spatial Analysis, Rutgers University
Radiometric resolution
•Sensitivity of the detector to differences in EMR signal
strength determines the smallest difference in brightness
value that can be distinguished
Dark Bright
Determined by the A-to-D quantization
6 bit = 0-63, 8 bit = 0-255, 10 bit = 0-1023
Center for Remote Sensing and Spatial Analysis, Rutgers University
Radiometric resolution
• Higher radiometric
resolution is especially
important for quantitative
applications such as sea-
surface temperature mapping
where the user wants to
distinguish small differences
in temperature
Center for Remote Sensing and Spatial Analysis, Rutgers University
Satellite remote sensing orbits give repeat coverage
• Geostationary Polar Sun-synchronous
• Constant view of hemisphere Covers entire Earth
700-900 km
35,800 km
Center for Remote Sensing and Spatial Analysis, Rutgers University
SPOT has steerable mirror to increase overpass frequency
Center for Remote Sensing and Spatial Analysis, Rutgers University
Change Detection
• The ability to
monitor change is
one of the
benefits of remote
sensing
• We can monitor
human and
natural changes
in the landscape
Center for Remote Sensing and Spatial Analysis, Rutgers University
Hurricane Andrew takes on Florida
Center for Remote Sensing and Spatial Analysis, Rutgers University
Spatial and temporal resolution
50 years Ground Surveys
As spatial 5 years Aerial Photography
resolution R
increases, the 3 years e Space photography
p
revisit e
time is also 28 days SPOT
a
increased, as t
Landsat-TM
are the 17 days
t
applications i NOAA AVHRR
that are 12 hours m
appropriate e Meteosats
and the cost 30 min.
.1 m 1m 5m 10-20m 30m 1 km 5 km
Ground resolution
Center for Remote Sensing and Spatial Analysis, Rutgers University
Many different systems. Which to choose?
Center for Remote Sensing and Spatial Analysis, Rutgers University
Different sensors and resolutions
sensor spatial spectral radiometric temporal
----------------------------------------------------------------------------------------------------------------
AVHRR 1.1 and 4 KM 4 or 5 bands 10 bit 12 hours
2400 Km .58-.68, .725-1.1, 3.55-3.93 (0-1023) (1 day, 1 night)
10.3-11.3, 11.5-12.5 (micrometers)
Landsat MSS 80 meters 4 bands 6 bit 16 days
185 Km .5-.6, .6-.7, .7-.8, .8-1.1 (0-63)
Landsat TM 30 meters 7 bands 8 bit 14 days
185 Km .45-.52, .52-.6, .63-.69, (0-255)
.76-.9, 1.55-1.75,
10.4-12.5, 2.08-2.3 um
SPOT 10m P / 20m X P -1 band X- 3 bands 8 bit 26 days
60 Km P - .51-.73 um (0-255) (2 out of 5)
X - .5-.59, .61-.68, .79-.89 um
IRS1 5.8 meters 1 band 6 bit 22 days
70 km .5-.75 (0-63)
IKONOS 1 meter 1 band 8 bit 3 days
11 km .45-.9 (0-255) (1.5 out of 3)
Center for Remote Sensing and Spatial Analysis, Rutgers University
Remote Sensing - Summary
• Remote sensing : acquiring and analyzing data using distant devices
recording electromagnetic energy
• Digital scanning devices or analog photography
• Airborne, terrestrial, marine, or orbiting platforms
• Can be passive, sensing existing radiation, or active, sensing radiation
bounced off the surface, as with radar.
• Remote sensing usually involves digital data, but also photography
Usually means both the capture and computerized image analysis of
data. Quantitative, easily incorporated into GIS
• Design of a remote sensing effort must clearly define information needs
and consider the 4 types of remote sensing resolution: radiometric,
spatial, spectral, temporal when considering the types of imagery to use
Center for Remote Sensing and Spatial Analysis, Rutgers University
