基于CamShift的压缩跟踪方法

针对压缩跟踪算法在目标旋转、尺度变化的情况下，容易出现跟踪漂移和跟踪失败的问题。本文采用改进的CamShift和压缩跟踪算法相结合的方法从不同角度对目标位置进行预测；然后，根据目标大小对跟踪窗口尺度进行自适应调整；最后，利用前后两帧图像的匹配程度自适应更新参数，提高算法的准确性。测试结果表明，本文算法不仅在目标外观变化的情况下具有较好的鲁棒性，而且在目标受到相似色干扰、遮挡和出现运动模糊时，仍能准确实时的跟踪目标。

Compressive tracking based on CamShift algorithm

when the tracking target is rotated or the scale of object has changed, compressive tracking always suffer the drift problem, even the tracking failed. In order to improve the accuracy of the algorithm, this paper combines the improved CamShift with Compressive Tracking to predict the target location from different angles. Then according to the size of target, the scale of the tracking window is adaptively changed. Finally, updating the parameters on the basis of the matching degree between the previous frame and this frame. The experimental results indicate that the algorithm is not only robust in the case of target appearance changed, but also it can real-time tracking the target under the condition of color interference, occlusion and motion blur.