CE 261: Lab #4
ImageVision Libraries Satellite Application Development, Part II

Spring 1996

Instructor: Dr. Craig M. Wittenbrink
Office: Applied Science Bldg. #309
Phone: (408) 459 4099
craig@cse.ucsc.edu

Due: Tuesday, June 11. Labs are to be demonstrated before the final exam (not necessarily on Tuesday). Notify the instructor to set up a time to demonstrate your lab.

Lab Environment: Computer Engineering/Computer and Information Sciences educational lab. Silicon Graphics Indy's and Indigos.

Location: Applied Sciences 213

Lab Hours: Any time, but you need a keycode

1.0 Description

In Lab #2, and Lab #3 you developed software that performed low level image and mid level image processing on satellite and radar images. In the fourth and final lab, you are to implement some more mid level vision operations and high level operations. Using the results of lab #3, segment out the clouds within the images of your choice. Use a transform based method of your choice to convert the image components to another representational basis (Hough, Fourier, KLT, etc). In the other representation space, use some means for classifying features in the original image. Any classifyer process can be used to decide between the clusters in your alternate representation space (line space, frequency space, principal component space, etc.)

Ideas of features to detect are ``internal waves'' and surface waves, cumulous clouds, buildings, or ships. There are very many targets which you can extract from the images.

2.0 Tasks

1. Preprocess, Via Sobel, thresholding, Laplacian, etc. The ImageVision libraries [1] provide a variety of approaches. Please see the graphical examples in the rear of the Image Vision libraries programming guide.
2. Segment, Is it easiest to see areas of discontinuities or areas of similarities? Divide the satellite data into clouds/ocean, coastline in any method you see fit. Work on a scheme to combine segmentations into reasonable objects.
3. Connected components, an algorithm for connected components is provided in Gonzalez and Woods, pages 42-43 [2]. A completely specified pseudo code connected component algorithm is provided in Haralick and Shapiro, pages 29-48 [3].
4. Decision vectors. Once you have done the connected components there are a variety of simple methods to separate them. How many pixels belong in a given component (area), what is the long axis of a component, the short axis, the ratio of long to short axes etc.
5. Transform to an alternate representational space.
6. Illustration of components (Required). Show in the image the separate components, either with number labels overlaid on the original image, pseudo colors with a key, or a similar scheme to illustrate the effectiveness of your algorithm. Illustrate classes, or targets, and not targets.
References

[1] ImageVision Library Programming Guide, by Jackie Neider and Eleanor Bassler, Document Number 007-1387-030, Silicon Graphics, 1993.

[2] Gonzalez and Woods, Digital Image Processing. Addison-Wesley, 1992

[3] Haralick and Shapiro, Computer and Robot Vision. Addison-Wesley, 1992.

Last Modified: 04:18pm PDT, May 24, 1996