QQ客服 官方微博 反馈留言

	Hand Detection and Gesture Recognition Exploit Motion Times Image in Complicate Scenarios
	Zhan Song; Hanxuan Yang; Yanguo Zhao; Feng Zheng
	2010
会议名称	6th International, Symposium on Visual Computing, ISVC 2010
英文摘要	Hand gesture recognition in complicate scenario is still a challenging problem in computer vision domain. In this paper, a novel hand gesturerecognition system is presented. To detect the exact hand target from complicate scenarios, the color and motion clues are used to obtain potentialhand regions. And then a method named Motion Times Image (MTI) is proposed to identify the optimal hand location. The R-transform descriptor is used to describe the hand shape features and an offline trained Support Vector Machine with Radial Basis Function kernels (RBF-SVM) is exploited to perform the hand gesture recognition task. Extensive experiments with different users under dynamic and complicate scenarios are conducted to show its high recognition accuracy and strong robustness.
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/172644/2857]
专题	深圳先进技术研究院_集成所
作者单位	2010
推荐引用方式 GB/T 7714	Zhan Song,Hanxuan Yang,Yanguo Zhao,et al. Hand Detection and Gesture Recognition Exploit Motion Times Image in Complicate Scenarios[C]. 见:6th International, Symposium on Visual Computing, ISVC 2010.

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。