<title>Iterative algorithm for designing binary phase-only composite filters</title>

Abstract

Abstract Use of the Magneto-Optic Spatial Light Modulator (MOSLM) in the filter plane of optical correlatorsis attractive because it is capable of high frame rates. This has led to the design of several differenttypes of Binary Phase-Only Filters (BPOFs) and Ternary Filters matched to a single reference. Tofurther increase the throughput of an optical correlator employing the MOSLM, attempts at designingBinary Phase-Only Composite Filters have been made. In this paper, we introduce a new iterativetechnique to design such Composite Filters in which we simultaneously maximize the peak sharpnessand minimize the deviation of the actual correlation peak intensities from their desired values. 1 Introduction Binary Phase-Only Filters have received much attention in the past ten years [1-9]. Most of this research has focused on the design of filters matched to a single reference image. These filters have beendesigned to optimize many different figures of merit. Among these are Signal-to-Noise Ratio, Peak-to-Correlation Energy and light efficiency. Now, the focus appears to be shifting to the design of Binary Phase-Only Composite Filters (BPOCFs) [10,11]. If we desire real-time implementation, then binary phase-onlycomposite filters, which are used to recognize many distortions of an image, should be considered sincethey require fewer filters.In this paper, we present a new algorithm for designing BPOCFs. The advantage of this new algorithmover the previous ones is that it allows us to not only approach the desired correlation peak value, butit also maximizes the peak sharpness produced by the filter. The remainder of this paper is organized asfollows. Section 2 introduces composite filters or synthetic discriminant functions (SDFs) [12,13] and setsup the notation we use. Section 3 develops the new algorithm for designing BPOCFs. Section 4 presentssome initial results and Section 5 gives our conclusions.