智能视频监控中目标跟踪的研究与实现

目标跟踪是智能监控领域的关键技术,本文主要是研究实现这一关键技术。提出一种融合运动目标的面积和矩特征进行匹配的算法,并结合运动估计算法对下一时刻目标最可能存在的方位进行预测,实现了对运动目标的跟踪。实验表明该算法能够准确跟踪运动目标,在一定程度上满足了实时性的要求。

Research and realization on tracking of object in intelligent video surveillance

The object tracking is the key technology of intelligent surveillance field,This paper is mainly to achieve this key technology.An method based on fusion objects area and torque characteristics for matching is put forward,and combining motion estimation algorithm to prediction the most possible objects location for the next moment,achieve the tracking of moving objects.It show that this algorithm can track the object exactly and can meet with requirements of real-time in certain degree through the experiments.