Maximum Average correlation Height filter for detecting distortion targets in cluttered environment using computational intelligence

Based on the computational intelligence and the principle of matched filter, the recognition problem of distortion target (rotation and scale etc.) in cluttered background scene is solved. The Maximum Average correlation Height (MACH) algorithm is adopted to design the matched filter, and control parameters of the filter are optimized. It is a very effective way to have high tolerance and to suppress clutter noise. As an experiment example, the target recognition of a tank in cluttered scene is presented. The experimental results show that a distorted tank can be recognized effectively after MACH filtering.