Improvement of pattern recognition with a heuristic design of correlation filters

A new method to design correlation filters for target recognition is presented. The method is based on a heuristic optimization of performance metrics. By the use of the heuristic algorithms the impulse response of a conventional composite filter is iteratively synthesized until a prespecified value of the used quality metric is obtained. Computer simulation results obtained with the proposed method are provided and compared with those of common correlation filters in terms of recognition performance measures in cluttered and noisy input scenes.