REAL, IMAGINARY, AND CLUTTER GABOR FILTER FUSION FOR DETECTION WITH REDUCED FALSE ALARMS

We fuse the output from our real Gabor function macro detection filter with the outputs from a clutter Gabor filter and an imaginary Gabor filter. This fusion reduces false alarm clutter (P(FA)), while retaining high probability of detection (P(D)). Initial test results are given for a large set of 34 scenes with 189 targets in six different classes with full 360-deg distortions and contrast differences in the presence of clutter. We find P(D) = 97.9% and P(FA) = 0.038%. All three new Gabor transform outputs are achieved on a correlator with different filter functions.