灰度共生矩阵在皮肤纹理检测中应用研究.pdfVIP

15 7 Vol. 15, No. 7 2010 7 Journal of Image and G raphics Ju ly, 2010 1) 2) 郭 航 霍宏涛 1) (, 100124) 2) ( , 102 23) W eb , , , , : TP391 : A : 100 -89 1( 2010) 07- 1074-05 Research on the Application of Gray Level Co-occurrenceM atrix for Skin Texture Detection 1) 2) GUO Hang , HUO Hongtao 1) (College of Comp uter Sc ience, B eij ing University of Technology, B eji ing 100124) 2) (D ep artm ent of Inf orm ation Security Eng ineering, ChineseP eop le. s Pu lic S ecurity Un iversity, B eij ing 102 23) Abstract Image. s texture feature detection has m any app lications in the pornograph ic mi ages monitoring and web mi ages filtering. This paper described themethods of texture analysis and the texture characteristics of the skin, then gave the texture detection m ethod based on gray level co-occurrencem atrix. Through the expermi ents, the paper inferred the key param eters of the grey level num ber and the offset value in the grey level co-occurrencem atrix com puting. The expermi ent conclus ion show s that contrast and energy are the best texture indexes wh ich could d istinguish the skin area and non-skin area. K eywor s pornographic mi ages, texture feature detection, gray level co-occurrencem atrix , , 0 , , , [ 1] , , ,