Plate Recognition Using Fuzzy Noise Removal and Opposition-based Micro-Differential Evolution

Automatic plate recognition of vehicles is of great importance in route management systems. Especially that such systems require real-time algorithms to perform the plate recognition task as soon as possible. In this paper, a new plate recognition system based on a fuzzy noise removal technique and the opposition-based micro-differential evolution (OMDE) algorithms is presented. Since population-based algorithms are mostly working based on a large population size, which results in a high computational cost, the micro-population-based algorithm require a very small population size and as a result, less computational cost. However, due to lack of diversity in such algorithms, the idea of opposition-based learning is utilized to enhance the diversity in the search procedure. By employing the template matching method, the digits and alphabets are easily extractable from plate image. Performance evaluation of the proposed system shows that this system is capable of identifying the vehicle’s plate with high quality in route management systems. Keywords—micro-differential evolution, plate recognition, opposition learning, template matching.