A system classification algorithm using neural networks
Rustom Mamlook, Wiley E. Thompson · 2002
A system classification algorithm to classify an image into targets and backgrounds is presented. The algorithm uses a double window filter in which the inner window surrounds the predicted target(s), and the outer window contains the predicted background(s). A solution to the problem of threshold selection is given. An example is given to illustrate the application of this algorithm for multi-target and multi-background classification.>