为有效地处理目标跟踪中的目标尺度变化、光照变化和局部被遮挡问题，采用分块跟踪策略，跟踪目标上的多个跟踪点，通过前后两帧图像间多对对应跟踪点之间的距离变化量确定目标尺度的变化量和位移量。在目标被遮挡的情况下，利用未受遮挡影响的跟踪点的位置信息和时空上下文信息计算目标跟踪点位置。采用高效而精确的STC（Spatial Temporal Context）跟踪算法跟踪单个跟踪点，保证算法的实时性和跟踪点定位的准确性。实验结果表明，提出的分块跟踪算法不但实时性较好，跟踪精度较高，具有抗光照变化和抗局部遮挡能力，而且算法原理简单，能够适应多种跟踪环境。

To solve the problems of object scale changing, illumination variation and partial object occlusion in target tracking, the object tracking strategy based on block method is adopted for tracking multiple points in the object box.The variation of object scale and displacement are confirmed according to the relationship between corresponding points change through two sequential pictures.Considering object occlusion, location of object tracking is calculated out by spatial temporal context information between tracking points and uninfluenced tracking points.To improve the real-time performance and locating accuracy of tracking points of the algorithm, the efficient Spatial Temporal Context（STC）method is applied to track every single point.It is demonstrated from experiment that this novel multi-block tracking method can track target accurately, and has the capacity of anti-partial occlusion and resistance to illumination changes.Meanwhile, this method is efficient, has good real-time performance, and is adapted to different tracking conditions.