Multi-angle Face Detection Based on Fuzzy Pattern Matching
基于模糊模式匹配的多角度人脸检测*

First, a pretreatment of color balancing and lighting compensation is executed on color image. In succession a skin color model in TSL chrominance space is used for segmenting regions where there may be faces, and then the face templates and fuzzy theory are utilized to search faces in these regions. Experimental results demonstrate this algorithm can satisfy the requirement of detecting multi-angle faces in complex environments.