An Effective Image Matching Algorithm Based on Rotation Independence
Traditional image matching algorithms has poor accuracy in image comparing,such as histogram intersection method.A new image matching algorithm based on the similarity comparison of irregular shape is presented in this paper,which divides the image into a number of irregular regions according to different colors,and extracts the boundary points of the irregular region to compose an irregular shape.The direction and distance is used to comparing the two irregular shapes if the rotation of the image is not considered,otherwise circular list is used to ignore the image rotation.It can be used widely.If two irregular shapes are similar,the two images are considered similar.Experiment proves that this method can effectively improve the image matching accuracy.
Image matching irregular shape blocking rotating-independence, circular list
Huang Xianying Chen Weiwei
School of computer science and technology Chongqing University of Technology Chongqing, China,400054
国际会议
太原
英文
354-357
2013-04-06（万方平台首次上网日期，不代表论文的发表时间）