结合局部差分的CamShift动目标跟踪算法

Cam Shift Tracking Algorithm Combining with Local Background Difference for Moving Target

Cam Shift算法具有计算量小、对目标形变适应能力强、执行效率高等优点,其在图像处理领域具有重要地位。但仅采用颜色直方图作为目标特征,忽略目标位置信息,会造成该算法在目标形态调整的过程中容易将颜色相似的背景纳入跟踪目标,从而发散甚至跟踪漂移。针对这一现象,提出了一种结合局部差分的Cam Shift跟踪算法,以进一步改善该算法的实际应用效果。改进后的算法首先结合Mean Shift迭代结果和扩展后统计区域的背景差分结果,综合判断移动物体的中心位置、长度、宽度等信息;然后通过此信息对运动目标形态变化的调整进行约束,使调整后的范围更加符合实际情况。试验结果显示,改进后的算法能够在复杂背景下准确地锁定跟踪目标,证明了结合局部背景差分技术的Cam Shift算法可以在增加较少计算量的情况下排除相似颜色的干扰,减少了跟踪过程中出现的不稳定现象,提高了算法的准确性和鲁棒性。

An important position of CamShift algorithm in the field of image processing is established with the advantages of small amount of computation,strong adaptability to target deformation and high execution efficiency. But only using the color histogram as the target feature,ignoring the position information of the target,may cause the background in similar color would be treated as the tracking target in the process of target form adjustment. Aiming at this phenomenon,the CamShift tracking algorithm combining with local difference is proposed to further improve the practical application effects of the algorithm. The improved algorithm firstly combines the results of MeanShift iteration and the background difference of the extended statistical region, to comprehensively judge the central position,length and width of the moving object. Then,with the help of this information,the adjustment for the deformation of the moving target is constrained, and the adjusted range is more in line with the actual situation. The experimental results show that the improved algorithm can accurately track the target under complex background, and prove that the CamShift algorithm combining with local background difference technology can eliminate interference of similar color,with adding less computation,and reduce the instability of tracking process,as well as improve the accuracy and robustness of the algorithm.