Affiliation:
1. Henan Institute of Science and Technology
Abstract
In color images, skin color is the important information on human face. This paper proposes a method to detect and locate human face rapidly based on skin color information and eye gradient. First, normalized RGB space is converted to HSV space; Secondly, the images are pretreated by smoothing and light compensation to overcome the uneven illumination changes, and then the defined skin color model is used to determine candidate regions of the human face, finally the human face is located accurately through eye localization based on gradient template. Experiments show that the method is fast and effective.
Publisher
Trans Tech Publications, Ltd.
Reference7 articles.
1. Huang Jianzhu, Feng Gui, Xia Lijuan. Fast face detection based on mixed skin color model. Image Processing and Multimedia Technology, Vol. 29, no. 18(2010), p.48.
2. Fu Guanhua, Chen Yiqiang, Liu Junfa. One method for automatic segmentation of human face. Application Research of Computers, Vol. 10, no. 25(2008), p.3060.
3. FY Shih, Y. Fu and K. Zhang. Multi-view face identification and pose estimation using B-spline interpolation. Information Sciences, Vol. 169(2005), p.189.
4. B. Moghaddam and A. Pentland. Probabilistec visual learning for object recognition. IEEE Trans. Pattern Analysis and Machine Intelligence, Vol. 19, no. 7(1997), p.696.
5. M, S, Lew. Information theoretic view-based and modular face detection. Pro. 2nd Int'1Conf. Automatic Face and Gesture Recognition, Vol. III(1996), p.198.