基于复杂特征融合的改进mean shift目标跟踪

Improved mean shift tracking algorithm based on complicated feature fusion

提出一种融合Gabor小波纹理特征与颜色特征的改进mean shift目标跟踪算法.首先,提取移动目标的颜色特征和纹理特征直方图;其次,基于mean shift算法定义融合相似度系数,对特征空间进行融合并得出目标中心位置;再次,通过定义特征自适应系数来融合基于颜色和纹理特征的目标位置;最后,对上述结果进行处理,得到目标最终位置.实验结果表明,该算法在跟踪目标存在变形、噪声、遮挡时能够得到比较理想的跟踪效果.

A novel improved mean shift algorithm based on texture and color features is proposed. Firstly, the moving object histograms of the color feature and texture feature are got respectively. Secondly, the fusion similarity coefficients are defined to fuse the different feature space, and the central location of moving object is calculated based on MS tracking. Thirdly, according to the color feature and texture feature, the object location is updated by using the feature adaptive coefficients. Finally, the above results are processed to get the final object location. Experimental results show that the proposed tracking algorithm exhibits good results in the presence of noise, deformation and occlusion.