一种新的基于自适应窗口的压缩跟踪算法

摘要

论文提出了一种新的自适应窗口调整的方法，首先提取每帧图像目标在跟踪框内的灰度值，计算其灰度均值，然后比较相邻两帧图像的灰度均值，通过其比值建立了一种自适应窗口尺度调整的数学模型，最后与压缩跟踪算法结合起来对尺度变化的目标进行跟踪。解决了压缩跟踪算法中跟踪窗口不能随目标大小而改变这一问题。从而使压缩跟踪算法在目标尺度发生变化时，及时调整跟踪窗口的尺度，从而提高跟踪精度。实验结果表明，改进后的算法比压缩跟踪算法具有更好的鲁棒性，跟踪性能得到提升。%This paper presents a new method of adaptive window method .Firstly the gray value in the tracking box of each frame is extracted ,the mean value is calculated .Then the gray average between two frames is compared ,a mathemati‐cal model is built for adjusting object size with an adaptive window by the ratio .Finally combined with compressive tracking algorithm ,the changing scale object is tracked .It copes with the tracking window which do not change with the changing ob‐ject during Compressive tracking algorithm .When the object size changes ,the scale of the tracking window is adjusted .So tracking accuracy is improved .The experimental results show that the improved algorithm has better robustness than the o‐riginal algorithm and the tracking performance has been improved .