Construction of Low Noise Optical Correlation Filters and Their Application to Target Identification Problems

摘要

Synthetic discriminant functions (SDF's) for optical matched filters have potential use for pattern recognition. However, these filters have been plagued with low signal-to-noise ratio (SNR); i.e., these filters have no trouble correlating very well with true targets, but very often give high (even major) correlations with false targets. In fact, numerical experiments by the author and others on realistic data sets show that the standard recipe for manufacturing SDF's gives filters with an SNR close to 1.00, even on a training set of imagery which has been edge-enhanced and energy-normalized. This document gives a new recipe for manufacturing SDG's. When applied to the data set of images. When tested against a randomly generated sequence of true targets in very cluttered backgrounds (true tanks in a junkyard of tank parts), this new filter so far has invariably picked out true target, whereas the filter manufactured with the standard recipe has given the major correlation to false targets approximately 25 percent of the time.