The Research of SURF Image Matching Method Based on Region and Feature Information

摘要

According to the SURF feature matching dense matching interpolation algorithm to realize the problems in the complex dynamic background image. The dynamic camera captured the same environment as the object. SURF image is proposed based on the combination of region and the feature information matching method. Firstly, image gray-scale normalization. There may be a reduction the brightness of light due to the translational response point of view. And then add several constraints in the process of extracting SURF feature points. Not only can speed up the matching process, but also can reduce the false match rate. Finally, in order to detect the feature points as the seed point for regional growth, the seed region division of the region by using graph matching. The test show that the algorithm can obtain dense disparity images of good. After the camera shooting of different sizes of image feature matching. Experiments show that, Compared with the SIFT algorithm and the SURF algorithm, the algorithm reduces the number of feature points and matching points, and improves the matching accuracy and the matching accuracy.