2.3 Chapter Overview – Beyond Whole Images
Chapter 13 introduces the Neural Net. The Neural Net matches a segment of an image with the other images; therefore, it is doing a different job from the three previous filters. Neural Net Matching is similar to the Unsupervised Filter, but instead of matching the whole image against another whole image, it matches a part of the images. Neural Net does require training, but the training is different from the other filters. The training is to set up the neural net for an Unsupervised Matching between an image segment and a whole image.
Chapter 14 introduces several examples using the Neural Net Filters.
Chapter 15 presents an example: finding Advertisement on TV program (digitized images).
Chapter 16 introduces Counting and Tracking. Counting counts the number of objects in an image, assuming there is no overlap between objects. Tracking finds the most obvious object in an image and tracks it from image frame to image frame.
Copyright (c) 1998 - 2006 Attrasoft, Inc. All rights reserved.