Authors: Gang Wang, Lizhu Liu, Yanqing Jin
Addresses: Department of Information Engineering, Zhengzhou Information Science and Technology Institute, Zhengzhou 450002, China. ' Department of Information Engineering, Zhengzhou Information Science and Technology Institute, Zhengzhou 450002, China. ' China National Digital Switching System Engineering and Technological Research Center, Zhengzhou 450002, China
Abstract: Fuzzy information theory which is based on fuzzy theory is a departure from information science. Higher performance could be obtained in pattern recognition with fuzzy sets theory. People combine fuzzy sets theory with pattern recognition and established fuzzy pattern recognition. This paper presents our work in the field of logo recognition by using fuzzy sets theory. Aim at the application requirement of logo recognition in image intelligent processing, a logo recognition algorithm based on membership degree and closeness degree of fuzzy sets is proposed. By fuzzy mapping, the logo|s gridding feature is transformed to the membership degree of fuzzy sets; then using closeness degree and closest principle of fuzzy sets to accomplish logo recognition. The algorithm significantly enhances the adaptability and anti-interference to poor quality images and effectively improves the flexible processing capability of logo recognition system. Experimental results show that the recognition rate of this logo recognition algorithm can reach as high as 94.5%.
Keywords: logo recognition; fuzzy sets; membership degree; closeness degree; closest principle; pattern recognition; fuzzy set theory; image processing.
International Journal of Modelling, Identification and Control, 2011 Vol.12 No.1/2, pp.83 - 87
Published online: 31 Dec 2010 *Full-text access for editors Access for subscribers Purchase this article Comment on this article