Image segmentation and target tracking based on meanshift algorithm

The kernel function is proposed for the analysis of meanshift algorithm, the research of meanshift algorithm mainly analyses the two common characteristics of meanshift algorithms based on kernel function histogram and probability distribution. Image segmentation based on meanshift is proposed for the extraction of target. The meanshift tracking algorithm based on the background difference suppression adopted, thus to give greater weight to the region with significant background difference around target and make sure the goal always in the tracking box. This thesis which is based on analyzing the characteristics of moving target tracking algorithm, uses the meanshift algorithm to accomplish moving target tracking. By comparing the Bhattacharyya coefficient of similarity function between target model and measurement model, the idea of meanshift is to find the position of target through multiple iterations drift, eventually reaching an ideal result, to complete the target tracking. The algorithm of target tracking based on meanshift in continuous video sequences is extensive. Therefore, target detection and tracking in continuous video sequence has a lot of practical and applied value. Experimental results show that the algorithm can help achieve fast and effective target tracking. It could be easily realized to get accurte tracking position for low-speed targets, but not suitable for tracking fast moving target. This algorithm can solve the problem in the tracking algorithm, that is when the pedestrian moving fast, it is easy to lose tracked target.