视频序列中基于肤色的人脸实时检测

Real-time Face Detection Based on Skin-color in Video Sequence

利用肤色在颜色空间的聚类特性来检测人脸,为了去除亮度对肤色的影响,增强检测的可靠性,采用了同时在RGB归一化颜色空间和HSV颜色空间中对肤色进行检测.在肤色检测之前采用了图像差分法先检测出运动空间,消除了背景对肤色检测的影响,也缩小了肤色搜索的范围.关于实际场景的实验表明,本文算法既能准确地检测出人脸,同时又能够达到实时效果.

This paper presents a method of face detection based on the clustering of human skin-color. In order to eliminate the effect of the luminance and to improve the exactness of detection, the skin-color is checked both in the RGB normalized color space and HSV color space. Before that, the motion area is extracted to eliminate the influence of the background, meanwhile, to minish the scale of the searching space. The experimental results show that the face area in the video sequence recorded in the real scene can be exactly detected in real time.