Researching On Filtering Of Object Image In Optical Correlation Detection

As a newly active discipline, optic information processing warms up the front edge of optics, electronics and computer science. It infiltrates in the field of science researching, technology development, industrial production, national defence and economy and acts as an even more important role. In this paper, we use real-time joint transform correlator to implement optic image detection, recognition and position. The process is fast and the accuracy of detection and recognition is very good.In the real-time joint transform correlator, for some unrecognized object images,which has low contrast, low brightness and cluttered scenes, the energy of the correlation images' correlation peaks is low, and correlation peaks are ambiguous and even sometimes could not be detected,so that the auto detection and recognition of objects could not be implemented. In this case, improved Laplace edge detection operator and the contrast stretching transformation were combined and applied to process the object image, the edge information was extracted from the image, and the contrast was increased too, then obtained more sharp correlation peaks and the experiment result is satisfying.