摘要
利用Lucas-Kanade光流法进行目标跟踪时,目标本身存在旋转、位移、缩放等情况,影响跟踪的准确性。因此提出一种新的算法。该算法先使用改进的PSO估算出一组参数,然后把更新出的参数送回到光流法,再进行一次更新,有效地计算出更合适的参数。实验结果表明,该算法能快速有效的对目标进行跟踪。
When tracking the target using Lucas-Kanade method for optical flow estimation, there are the impacts on the tracking accuracy caused by the target rotation, translation and scaling. Therefore, we propose a new algorithm. The algorithm first uses the improved PSO to estimate a set of parameters, then it sends back the parameters formed from updating to optical flow method and updates once again for the effective calculation of the more appropriate parameters. Simulation results show that this algorithm can rapidly and effectively track the moving objects.
出处
《计算机应用与软件》
CSCD
北大核心
2014年第7期217-220,共4页
Computer Applications and Software
基金
广东省自然科学基金项目(S2011010003442)
