Ferroelectric liquid crystal over VLSI silicon spatial light modulators have been successfully employed as input and filter devices in optical correlator architectures for image processing and target recognition. A considerable limitation of these systems is their difficulty in extracting the desired target from clutter, particularly where this clutter has similar characteristics to the target. We present a novel set of image processing techniques for improving the target to clutter ratio, thereby simplifying the task of achieving correct target recognition. A stereo-pair of images captured from the input scene is enhanced using morphological image processing functions implemented as convolutions with an output threshold. Stereo pairs allow extraction of regions of interest based upon range. These regions are automatically extracted from the enhanced images ready for target recognition via a conventional correlation function. The clutter reduction algorithm does not require a priori knowledge of the target and so is a robust method for target recognition. All of the functions used can be expressed in terms of correlations and convolutions with an output threshold, allowing implementation on a correlator processing architecture. Initial experimental results from a Vanderlugt 4f optical correlator utilizing SLMs 256 by 256 pixel at the input and filter planes are presented. These results are compared to those from computer simulations and the performance of the optical system is assessed.
Replication in the output plane of an optical correlator, due to pixelation of the Fourier plane filter, can lead to false correlation signals. This paper suggests randomization of the pixel positions of the Fourier plane filter as a solution, and demonstrates its effectiveness through computational simulations and optical correlation experiments using a custom made SLM.