Face locating and tracking method based on chroma transform in color images

摘要

This paper has investigated the color facts in color images, obtained the relationship between chroma and color coordinates, and established a type of coordinate transform which is able to improve chroma of the skin and lip. With these coordinates, a new method of human face locating and tracking based on skin chroma and lip chroma transform in color images has been presented. It is effective and robust to detect the position of objects, not impacted by the pose of object and complexity of background. It is able to overcome the influence of illumination variance or different camera devices by on-line learning. This method is fast and stable. This algorithm is prerequisite for most human face perception system. By this approach, a feasible, reliable and real-time property algorithm to work out locating and tracking problem in real system is obtained.