Image Compression and Reconstruction based on Fuzzy Relation and Soft Computing Technology
Kaoru Hirota, Hajime Nobuhara, Kazuhiko Kawamoto, and Shin’ichi Yoshida
Department of Computational Intelligence and Systems Science, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8502, Japan
Received:June 25, 2003Accepted:September 9, 2003Published:January 20, 2004
Keywords:image compression and reconstruction, fuzzy relation, soft computing, linear quantization
A fast image reconstruction method for Image Compression method based on Fuzzy relational equation (ICF) and soft computing is proposed. In experiments using 20 images (Standard Image DataBAse), the decrease in image reconstruction time to 1/132.02 and 1/382.29 are obtained when the compression rate is 0.0156 and 0.0625, respectively, and the proposed method outperforms the conventional one in the Peak Signal to Noise Ratio (PSNR). ICF using nonuniform coders over YUV color space is proposed in order to achieve effective compression. Linear quantization of compressed image data is introduced in order to improve the compression rate. Through experiments using 100 typical images (Corel Gallery, Arizona Directory), PSNR increases at 7.9-14.1% compared with the conventional method under the condition that compression rates are 0.0234-0.0938.
Cite this article as:K. Hirota, H. Nobuhara, K. Kawamoto, and S. Yoshida, “Image Compression and Reconstruction based on Fuzzy Relation and Soft Computing Technology,” J. Adv. Comput. Intell. Intell. Inform., Vol.8 No.1, pp. 72-80, 2004.Data files: