Shape description matrix and its applications to color-image retrieval and recognition  

作　　者：WANG Yi,ZHAI Hongchen,LIANG Yanmei,ZHANG Siyuan & MU Guoguang Institute of Modern Optics,Nankai University,Tianjin 300071,China Key Laboratory of Information Sci-ence and Technology MOE,Tianjin 300071,China 

出　　处：《Science China(Technological Sciences)》2004年第2期159-165,共7页中国科学（技术科学英文版）

基　　金：the National Natural Science Foundation of China (Grant No. 60177004);the Natural Science Foundation of Tianjin; the Startup Foundation of Personnel Division of Nankai University.

摘　　要：Shape description matrix (SDM) describing regional pixel distribution characteristics of images is defined, with both radial and angular distribution messages taken into account to offer a comprehensive and exact description of the shape characteristics even upon a blurred edge. Using this SDM, a general color histogram (GCH) is developed. Both theoretical and experimental results show that the GCH can be applied to color-image retrievals and recognitions with a high efficiency.Shape description matrix (SDM) describing regional pixel distribution characteristics of images is defined, with both radial and angular distribution messages taken into account to offer a comprehensive and exact description of the shape characteristics even upon a blurred edge. Using this SDM, a general color histogram (GCH) is developed. Both theoretical and experimental results show that the GCH can be applied to color-image retrievals and recognitions with a high efficiency.

关 键 词：SDM  similarity degree  MATCH degree  general COLOR histogram. 

分 类 号：O439[机械工程—光学工程]