遮挡情况下的多人跟踪方法研究

Research on Multi-Human Tracking under Occlusion

针对步态识别中遮挡情况下的多人跟踪问题,提出了一种以运动人体的质心代替运动人体的跟踪方法。该方法利用人体质心表示运动的人体,给出了质心的计算方法;并且在多人遮挡的情况下,采用灰色预测模型对遮挡时的运动人体质心进行预测。利用该方法在自建的人体运动步态数据库上进行了实验,结果表明,该方法能够较好地解决遮挡情况下的多人跟踪问题。

With an aim at the multi-human tracking problem in the gait recognition under the occlusion, this paper suggests a kind of tracking method with the centroid of moving human body to replace the moving human body. This method is to use the human body eentroid to express the human body in motion and offers the calculation method of centroids. Also, in the case of the multi-human occlusion, the grey prediction model is adopted to prodict the centroids of human body in motion. The experiments are carried out on the secf-established guit database of human bodies in motion by means of this method. The results indicate that this method is able to solve the multi-human tracking problem in the case of occlusions better and quickly.