Automated processing in food processing industry

Document Sample
Automated processing in food processing industry Powered By Docstoc
					   Automated processing in food processing industry
                                                                                                             application story
                                                                                                             food processing industry

»Off with their heads!« Using an image processing system to cut carrots
Is it possible to process carrots automatically using image proces-
sing systems and to determine the point at which the green of the
carrot should be cut off? A feasibility study was carried out at Iris
Vision in Holland in order to answer this question. It was found that
the hardware and software available are able to complete this task
as quickly and accurately as required.

Vegetables are still harvested and processed largely manually. Not
only are these activities generally monotonous and in some cases
bad for the health of those involved; the associated labor costs also
have a considerable effect on the prices of the end products. The
Dutch distributor of the Common Vision Blox software system, Iris
Vision, a company based in Etten-Leur, therefore decided to examine
the extent to which image processing systems can be deployed in
this context.                                                            The application's graphical user interface

A feasibility study was carried out to examine whether image pro-
cessing systems were able to determine the point at which the car-
rot green should be cut from the carrot during automatic carrot
processing. If too little is cut off, some of the unwanted green tops
of the carrots remain. If too much is cut off, on the other hand, this
has a negative effect on the productivity of the process.                Different carrot types

The specifications
In order to create conditions that were as realistic as possible, a      The following steps are thus involved in determining the point at
conveyor belt was used to transport the carrots to the image pro-        which to cut the carrot:
cessing system. Pre-washed carrots between 150 mm and 300 mm             1. Ascertain the carrot's angle of rotation.
in length were placed on this at arbitrary angles and transported        2. Trace the thickness of the carrot along the longitudinal axis and
past the camera. Further conditions were that not more than                 find the thickest point.
10 percent of the total length of the carrot should be cut off and       3. Measure how rapidly the carrot decreases in thickness on both
that the part cut off should be less than 3 cm in length. This was          sides of the thickest point.
ensured by an upstream process. 20 carrots per second had to be          4. The point at which the reduction in thickness is greatest is the
processed.                                                                  approximate point at which the carrot is to be cut.
                                                                         5. Calculate the exact point at which the carrot is to be cut by
»Since the carrots can be lying in any position on the conveyor belt,       means of an accuracy adjustment procedure.
the first task is to ascertain their angle of rotation,« explains
Dietmar Serbée, the head of distribution at Iris Vision, who was in         facts
charge of carrying out the study. »The next step is to determine the
point at which the green part of the carrot is to be cut off.«           Industrial sector:       Food processing industry
                                                                         Task:                    Automated processing of carrots
The top of the carrot is the end at which the carrot is at its thickes   Hardware:                  Cameras M10 BX (JAI)
perpendicular to the longitudinal axis. From the point of maximum                                   Frame grabber PcVision (DALSA Coreco)
thickness, the carrot decreases in thickness much more quickly           Software:                  Common Vision Blox Image Manager and
toward the green end of the carrot than toward the root. The algo-                                  CVB tools Edge, Blob (STEMMER IMAGING)
rithm therefore searches for the cutting point on the green side of      Distributor:             Iris Vision, The Netherlands
the thickest point.
   Automated processing in food processing industry

Achieving the objective with the right software
»We solved this problem on the software side by using the tools           In the case of carrot inspection, CVB Blob begins by identifying the
Blob and Edge from the Common Vision Blox image processing                perimeter of the carrot. It can then calculate the moments of iner-
toolbox,« explains Serbée. This quasi-operating system for image          tia and use these to calculate the angle of rotation. »A carrot's mini-
processing was developed by the Puchheim-based German compa-              mum moment of inertia will generally be considerably lower than
ny STEMMER IMAGING GmbH in order to provide a standardized,               the maximum moment of inertia, which is essentially the carrot's
open platform allowing the rapid and flexible development of              longitudinal axis,« says Serbée. "CVB Blob uses this to calculate the
image processing applications. There are now over 30 Common               carrot's angle of rotation."
Vision Blox tools, and these can be used to solve virtually any indu-
strial image processing problem.                                          The Common Vision Blox tool Edge allows the edges of an object to
                                                                          be determined on the basis of the gray-scale values of the object
CVB Blob is able to calculate the shape of an object consisting of        and those of the background. A standard algorithm used by this
contiguous pixels. It calculates not just the object's centroid (center   software detects two opposite edges and calculates the distance
of mass), bounding box (the smallest rectangle that can surround          between these points with a level of accuracy that can be in the
the object), perimeter and area, but also its minimum and maximum         sub-pixel range, if required. These distances are saved, thus allo-
moments of inertia.The relationship between these two moments of          wing the maximum diameter of the carrot (its top) to be ascer-
inertia determines whether or not the object has an orientation. If       tained and the decrease in thickness from this point to be calcula-
the minimum and maximum moments of inertia are identical, as              ted. CVB Edge thus carries out steps 2 and 3.
they are for a circle, the object does not have an orientation.

Industrial implementation
An M10 BX camera from the Danish manufacturer JAI (a CCIR indu-           the length of the carrot. Within this window, CVB Edge looks for
strial camera with a resolution of 768 x 572 pixels) and a PcVision       pairs of edges representing the transitions from black to white and
PCI frame grabber card from DALSA Coreco were used for image              vice versa and saves the distances between these points. A further
acquisition in the feasibility study. »The images obtained in this        ROI is defined either side of the line representing the greatest
way are displayed as gray-scale images in a window in the upper-          distance between points on the opposite edges. In this ROI the
left corner of the graphical user interface,« explains Serbée. »The       software then finds out on which side of this line the thickness of
image is then converted into a black-and-white image with a               the carrot decreases more rapidly. This is the point, subsequently
»white« carrot on the basis of a predefined gray-scale threshold          adjusted by a defined distance, at which the blade later cuts the
value, which depends on the type of lighting used. In this step BLOB      green part of the carrot off. This adjustment provides a safety mar-
also ascertains the position of the carrot by calculating its centroid,   gin, thus ensuring that none of the green part of the carrot
and it displays its bounding box in the window in the upper-right         remains.
corner of the GUI.«
                                                                          The accuracy of this procedure depends on the width of the ROI in
CVB Blob then calculates the object's moments of inertia so as to         which the software calculates the side on which the thickness of
determine the carrot's angle of rotation on the basis of the maxi-        the carrot decreases more rapidly and on the size of the incre-
mum moment of inertia. An image of the carrot aligned horizon-            ments used to set the ROIs. The narrower the ROIs are, and the
tally is displayed in the lower-right corner of the GUI. In the next      smaller the distances between them, the greater is the accuracy.
step, a small region of interest (ROI) is defined perpendicular to the    A higher level of accuracy does, of course, inevitably slow the appli-
longitudinal axis. This can be set in adjustable increments along         cation down.
   Automated processing in food processing industry

Time-consuming calculation of orientation
The tests for checking the correct functioning of the algorithm                  rotation in degrees                  0°          45°          135°
were carried out with two different systems:                                                                       time [ms]   time [ms]     time [ms]
  System A: Intel Pentium Pro 200 MHz, 64 MB RAM                                 PENTIUM PRO 200
  System B: Intel Pentium III 500 MHz, 128 MB RAM
                                                                                 Blob and moments of inertia         896         885           896
Windows NT Workstation 4.0 with Service Pack 5 was running on
                                                                                 Edge detection                      142         207           174
both systems. Common Vision Blox Version 1.4 and Microsoft Visual
Basic Version 5.0 were used.                                                     Detection of cutting point           18          42            35

                                                                                 Total time                          1056        1134          1105
Table 1 shows the benchmarks for the different steps. They were
                                                                                 PENTIUM III
carried out with a resolution of 25 pixels per ROI. This table indi-
                                                                                 Blob and moments of inertia         395         390           451
cates that neither of the two systems will reach the required speed
of 20 carrots a second without further optimization. In addition,                Edge detection                       44         102            82
the measured values indicate that around 85 percent of the pro-                  Detection of cutting point           7           17            14
cessing time taken is required to determine the carrot's angle of
                                                                                 Total time                          446         509           547
rotation (i.e. to find the object and calculate its moments of inertia).
                                                                                Table 1: Benchmarks for the different steps
There are a number of options available for optimizing these values
and achieving the specified speed. If the carrots can be inspected in
a known position rather than at arbitrary angles of rotation, the               mates that the processing time can be reduced by a further 10 to

                                                                                                                                                           PP-IRIS1e-10/2005 . Subject to technical change without notice. No liability is accepted for errors which may be contained in this document.
processing time is reduced to less than 50 ms because the proces-               20 percent if the system is implemented in C++.
sor-intensive algorithms for calculating the areas and moments of
inertia are not required. This makes it possible to inspect 20 carrots          The required level of accuracy was reached in all the tests. To furt-
a second. There is also potential in the software itself for further            her increase the speed to meet more stringent requirements, it
increasing the speed: »The demonstration program is written in                  would also be possible to increase the size of the ROI and the size
Visual Basic and not optimized for speed,« explains Serbée. He esti-            of the increments. This would also lead to better performance.

The feasibility study shows that it is possible to use industrial               lopment there. »We are confident that this system will soon be
image processing to determine the point at which the green part                 able to demonstrate its effectiveness under industrial conditions,«
of the carrot should be cut off. The demonstration and evaluation               emphasizes Dietmar Serbée.
program developed on the basis of Common Vision Blox worked
robustly without optimization measures and achieved the required                    our partner iris vision
speed of 20 carrot inspections a second without software optimi-
zation, provided the objects are fed to the image processing station             Founded in 1996, IRIS VISION in the Netherlands offers a complete
at a specified angle of rotation. This would be easy to do using                 range of machine vision and image processing products from major
mechanical devices.                                                              manufacturers. IRIS´s expertise is the integration of all necessary
                                                                                 vision components for a wide variety of image processing and machi-
                                                                                 ne vision applications.
The hardware components (JAI M10 camera and PcVision frame
grabber from DALSA Coreco) were appropriate for the purpose and                  Ranging from the beginning of the data chain with lighting and
were able to provide the required resolution for reaching the requi-             optics, through the versatility of the acquisitions and processing to
red level of accuracy. Both components are supported by Common                   the result of the application. IRIS offers a range of solutions, simple
Vision Blox.                                                                     frame grabbers, extensive pipeline processors, host based processing
                                                                                 and hardware processing.
The demonstration system presented is currently still with the
                                                                                 More information:
customer for evaluation purposes and is undergoing further deve-
Germany:     Gutenbergstr. 9-11 . D-82178 Puchheim . Phone + 49 (0)89 / 80 90 2-0 . Fax + 49 (0)89 / 80 90 2-116
Switzerland: Rietbrunnen 48 . CH-8808 Pfäffikon . Phone + 41 (0)55 415 90 90 . Fax + 41 (0)55 415 90 91