Real-time Face Detection Based on Skin Color Model

摘要

In many applications it is desirable to implement a face detection system in real time, such as video surveillance, human computer interface, face recognition. In this paper, we propose a scheme based on the skin color model for fast face detection under non-constrained scene condition. First we build a skin color model that combines the skin color model in HSV color space with KL skin color filter. Our method detects skin regions over the entire image and then generates face candidates based on the spatial arrangement of these skin patches. Then we search for the mouths in all the candidate human face regions by means of mouth detector to verify each face candidate. The experimental results show that the proposed method is not only fast and efficiency, but also robust to head rotation to some extend.