基于均值漂移算法的人脸自适应跟踪

Adaptive human face tracking based on mean shift algorithm

采用自适应人脸方向模板和YCbCr自适应肤色模型,提出了一种新的基于均值漂移算法的自适应人脸跟踪方法。与传统的均值漂移跟踪方法相比,当人脸倾斜时或光线变化时,该方法能更精确地描绘出人脸位置。实验结果表明,在基本上不增加计算量的情况下,该方法能对人脸的倾斜和光线的变化进行很好地自适应跟踪。

By using an adaptive facial orientation template and an adaptive YCbCr skin color model, a novel adaptive human face tracking method was proposed on the basis of mean shift algorithm. Compared with the traditional mean shift tracking method, the method can describe the face position more accurately when facial orientation changes or lighting changes. The experimental results show that the method can adapt to the facial orientation and lighting variations without increasing calculated amount.