QQ客服 官方微博 反馈留言

	An algorithm of pornographic image detection
	Hong Zhu ; Shuming Zhou ; Jianying Wang ; Zhongke Yin
	2007
会议名称	Proceedings of the 4th International Conference on Image and Graphics, ICIG 2007
英文摘要	Real-time detection of pornographic images can effectively prevent pornographic images from spreading on the Internet. The methods ofdetection mainly focus on the identification of skin region. To overcome the disadvantages of traditional algorithms, in this paper, we propose a skin model based on the combination of YIQ, YUV, and HSV. In the step of the pre-dealing, we use white balance algorithm to achieve better skin area. Then, texture model based on Gray Level Co-Matrix (GLCM) and geometric structure of human beings are used to decrease the disruptions ofthe background region similar with the skin area. The features which extracted from the last images dealt by color and texture model are input into Support Vector Machines (SVM), through which the pornographic images and assorted control images are classified successfully. An experiment using 400 pornographic images and 400 assorted control images is carried out. Experimental results show that the algorithm of skin texture-based and feather-based detection has perfect performances, and the detection rate can reach 88.89%
收录类别	EI
语种	英语
内容类型	会议论文
源URL	[http://ir.siat.ac.cn:8080/handle/172644/2135]
专题	深圳先进技术研究院_其他
推荐引用方式 GB/T 7714	Hong Zhu,Shuming Zhou,Jianying Wang,et al. An algorithm of pornographic image detection[C]. 见:Proceedings of the 4th International Conference on Image and Graphics, ICIG 2007.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。