一种基于SIFT和KLT相结合的特征点跟踪方法研究

A Feature Point Tracking Method Based on The Combination of SIFT Algorithm and KLT Matching Algorithm

在目标发生明显姿态和大小变化条件下,为了利用基于特征点的跟踪算法实现对目标可靠、稳定跟踪,提出了一种SIFT算法和KLT(Kanade-Lucas-Tomasi)匹配算法相结合的特征点跟踪方法。通过对SIFT算法进行优化,使得到特征点分布相对均匀,同时不存在聚集现象;通过对KLT匹配算法进行分层迭代设计,提高了目标作快速运动时的匹配精度;最后根据特征点匹配结果,结合Greedy算法得到目标的准确位置。实验结果表明:该算法能够很好地适应目标姿态和大小的变化,实现对结构复杂目标的稳定跟踪;比KLT跟踪算法具有更好的鲁棒性和稳定性,能得到更加准确的目标位置。

In order to track target stably and accurately by the feature point tracking algorithm when the target changes in size and attitude,a new feature points tracking algorithm is proposed based on the combination of SIFT algorithm and KLT matching algorithm.Through the optimization of the SIFT algorithm,the feature points distribute relative uniform and there is no over-clustering of feature points.The matching accuracy is improved a lot when the target moves fast by the hierarchically iterative matching of the KLT matching algorithm.According to the results of feature points matching,the exact location of target is obtained with the combination of the Greedy Algorithm.Experimental results show that the proposed algorithm can adapt to objective changes in attitude and size and it can track target with complex structure stably.It has stronger robustness and reliability and get more accurate location of target than KLT tracking algorithm.