基于Canny-Harris-SIFT算法的立体匹配

摘要

Aiming at the problems of large calculating amount and high complexity in scale invariant feature transform algorithm,a new stereo matching method was proposed based on Canny-Harris-SIFT algorithm.In the matching method,Canny edge detection was combined with the matching methods based on Harris comer and SIFT feature.The feature points of images were detected by using Harris operator,the main orientation for each feature point was calculated,the feature point descriptors were generated,so as to improve the efficiency of feature matching.The experimental results show that using this algorithm,the complexity of feature extraction and matching is reduced,the real-time performance of SIFT algorithm is improved and the correct image matching rate is ensured.%针对SIFT算法存在计算量大、复杂度高的问题,提出一种基于Canny-Harris-SIFT算法的立体匹配新方法.该算法将Canny边缘检测、基于Harris角点的匹配方法和基于SIFT特征的匹配方法有机结合,采用Harris提取特征点,为每个特征点定义主方向和生成特征向量描述子,从而提高特征匹配效率.实验结果表明:该算法降低了特征匹配的复杂度,提高了算法的实时性和图像的正确匹配率.