To improve the tracking robustness of the traditional particle filter target tracking algorithm under the condition of illumination change and occlusion,particle filter target tracking algorithm based on dual-feature is proposed.Firstly,the paper presents a novel Sobel Med-value Binary Pattern (SMBP) feature description operator with light resistance and edge information,and establish texture histogram for video each frame.Color histogram is obtained by linear weighting of the hue kernel function histogram and saturation kemel function histogram in HSV color model.On this basis,for each frame of color histogram and texture histogram are normalized respectively,and introduced into the particle filter algorithm to calculate the particle weight of target characteristics,the final location of the current frame target is determined through the weighted criteria again.The experimental results show that the proposed algorithm can improve the accuracy of target tracking effectively under the condition of illumination change and occlusion.
NETL
NSTL
万方数据
维普
