A real-time white light pseudocolor encoding technique for spatial frequency and density encodings is presented. In spatial frequency color coding, the encoding is accomplished by spatial filtering of the color signal spectra, while in density pseudocoloring, the encoding consists of contrast reversal of a color object image. The technique is simple, versatile, and economical to operate, which may offer some practical applications. Because the encoding colors are primarily derived from a white light source, the annoying coherent artifact noise can be substantially reduced. Since the encoding is obtained with a broad spatial band of the signal spectra, this technique offers no apparent resolution loss. We stress that this real-time white light pseudocolor encoding technique may offer several major advantages that previous techniques have offered. Experimental demonstrations of this pseudocolor encoding technique are also provided.

A coherent optical information parallel processing technique using a multidiffraction grating is presented. We have shown that this parallel processing technique is capable of performing multichannel spatial filtering in the spatial frequency plane. The number of parallel processing channels is about 2N times the number of processing channels of the conventional coherent optical processor. However, in practice, the number of processing channels is limited by the number of multiplex gratings that can be synthesized and the available power of the light source. Extension of the monochromatic parallel processing system to polychromatic parallel processing is also presented. It can be shown that the number of processing channels of the polychromatic parallel processor is higher than the monochromatic parallel processor. Extension of this parallel processing concept to multisignal parallel processing is also discussed. However, the multisignal processing technique is not a real-time method because it requires an initial encoding step. Nevertheless, this disadvantage may be alleviated with ingenious design of the encoding system. We also assert that this parallel processing technique may be applied to white light parallel processing.

An optimization technique for the optical correlation detection process is proposed. The noise correlation output is suppressed by prewhitening its spectrum. Since the optimization operates on the noise instead of the signal spectrum, the system tolerance for size and orientation variation is not severely affected. Experimental illustrations are also presented.

No abstract

Correlation filter functions implemented with biphase-only modulators represent the state of the art in real-time programmable optical correlators for target recognition. High optical throughput, however, results at the expense of distortion-invariant performance. We present results of computer simulations and experiments using biphase-only modulation for the filter functions, which are designed to approximate the performance of correlation filters having both biphase and biamplitude modulation. Specifically, we investigated the effect of clutter in the input scene on correlation performance using these virtual amplitude filters and compared them with those for classical method filters and previous biphase-only filter designs.