国产欧美精品一区二区,中文字幕专区在线亚洲,国产精品美女网站在线观看,艾秋果冻传媒2021精品,在线免费一区二区,久久久久久青草大香综合精品,日韩美aaa特级毛片,欧美成人精品午夜免费影视

ABSTRACT:The target tracking algorithm based on correlation filtering is a common visual tracking method, which uses the feature information of the target to track; In the tracking process, the target position is determined by calculating the correlation between the target template and the candidate area in the current frame image; Through the introduction of the first MOSSE algorithm combining the concept of correlation filtering with target tracking technology, three improved correlation filtering tracking algorithms based on this algorithm are introduced: KCF algorithm, DSST algorithm and BACF algorithm; And conduct simulation experiments on MATLAB platform based on the OTB100 data set; Under the one-time evaluation standard, the average success rate and average accuracy of BACF algorithm are the highest 64.5% and 80.4% respectively; Under the spatial robustness evaluation standard, the average success rate and average accuracy of BACF algorithm are the highest 58.2% and 78.6% respectively; Under the time robustness evaluation standard, the average success rate and average accuracy of BACF algorithm are the highest 65.8% and 85.1% respectively; Therefore, BACF algorithm has the best tracking performance, while KCF algorithm achieves the highest tracking speed of 154.36 frame rate.