Image Matching Algorithm Combining SURF Feature Point and DAISY Descriptor
Na Luo · 2014
Image matching is a basic technique in the research of computer vision,and local feature based image matching methods are becoming increasingly popular in this field.To solve the classical SURF algorithm's poor performance on the rotation invariance,this paper proposed a new matching algorithm combining the SURF feature point and DAISY descriptor.We proposed a main orientation distribution method which is more suitable for DAISY descriptor,so that a new descriptor can be obtained via rotating by the main orientation.Our algorithm effectively improves the matching ability of the classical SURF algorithm on the rotation invariance,only employing a little more computational burden.The experimental results demonstrate that our algorithm is more robust than classical methods when the image blurs,illumination,JPEG compression ratio or the viewpoint changes.