基于卡尔曼滤波的人脸跟踪算法

针对实时视频监控领域中传统的Camshift算法不能自动跟踪人脸和容易受到肤色相近遮挡等问题,采用Ad-aboost算法实现了人脸的自动检测,同时对于跟踪丢失等情形,通过卡尔曼预测对跟踪偏差进行实时改进。实验表明跟踪的准确性有较大提高,具有较好的实时性;在相近肤色遮挡时仍能实现正确跟踪,并对侧脸也有较好的效果;算法具有较好的鲁棒性。

Auto face tracking base on kalman filter

In the field of real-time video monitoring,the classic Camshift algorithm cannot automatically track the human face and is easily subject to color interference covered and other issues,the paper used Adaboost algorithm to detects the human face automatically,and the tracking error in real time is improved through kalman prediction for the case that the track is lost.Experimental results show that the accuracy of track can be improved greatly,there has better real-time even when the block in the similar color to achieve the right track,and side faces also have a good effect,also the algorithm has better robustness.