中文摘要：

针对复杂图像的快速匹配问题,提出一种新的基于Shi-Tomasi角点检测与SIFT算法的高精度快速匹配方法。该方法充分利用图像的角点特征、灰度和位置信息,采用SIFT算法中的特征描述方法进行图像特征描述,并用Ransac算法对匹配点进行校正和消除错误匹配,提高计算速度和可靠性。实验结果验证了该算法对于存在较大色差、形变等图像可实现精确快速匹配,其精度和速度都优于传统的匹配算法。

英文摘要：

For fast matching of complicated images, a new fast matching algorithm with high accuracy based on Shi-Tomasi comer detection and SIFT algorithm is presented in this paper. It makes full use of graphic point feature, gray-level and location information, adopts the feature description method in SIFT algorithm to carry out image＇s feature description. Moreover, RANSAC method is used to correct the matching points and to remove false matching. These make it faster and more reliable in computation. Experimental results show that the algorithm can realise accurate and fast matching of the images with higher chromatic aberration or deformation, its precision and running speed all outperform the traditional matching algorithms.