Fuzzy Color Computing Based on Optical Logical Architecture

Abstract

This paper is dedicated to the development of intelligent techniques of optical computing for real-time decision support systems (DSS) with a large array of fuzzy input data. Currently offered as an alternative to the binary systems, fuzzy optical computing devices are very complex. To increase the efficiency of logical systems in the formation and processing of an array of input fuzzy data, we propose to use a light emitter of a certain color as a fuzzy set (FS) that is a carrier of logical information. This allows building of logical solutions and conclusions based on the additive and subtractive conversion of light radiation by appropriate color filters, measuring illumination in optical channels, and switching light emitters. The main algorithmic logical procedures based on optical gates that perform the basic logical operations of disjunction and conjunction, negation, and search for a new solution are considered. To solve a problem with a large amount of fuzzy data, we propose the architecture of networks of logical structures and inference procedures. Also, we implement optimization of the optical circuit solutions to increase the reliability of decisions and estimates. When describing color sets using a fuzzy sequence, we propose to implement a non-linear rank scale. This increases the accuracy of solutions, as shown by the decrease in root-means-square estimation error.