A novel approach for multi-pose face detection by using skin color and FloatBoost

Multi-pose face detection plays an important role in the automatic human face recognition system. The YCbCr color space is used first for searching the skin color regions, then, raster scanning is adopted to obtain the four-connected components, identify the center of each labled block, and search and clip the potential face regions containing the triangular feature, finally, the classifier based on FloatBoost is designed to determine whether a potential face region contains a face. Experimental results demonstrate that the proposed method performs better than previous methods in terms of speed and ability to handle multi-pose face conditions.