An image matching method for arbitrary rotation
In the process of image matching,it is unavoidable to generate a variety of distortions and angle rotation is a common form of image distortion. So antirotation matching has become one of the most important aspects of image matching algorithm research. According to the thesis,it obtains the relative rotation angle by extracting the rotation invariant area of the image and calculati.Then it achieves the image matching with the normalized product correlation coefficient as criteria for the similarity measurement. The result of the experiment shows that the method can be resistant to arbitrary rotation and has a higher accuracy than the traditional circular projection matching algorithm.
image matching anti-rotation circular projection algorithm normalized product correlation
Wenjian Chen Hong Cheng Wenbang Sun Guang Yu
Department of Specialty,Aviation University of Air Force,Changchun,China
国际会议
西安
英文
992-996
2011-12-23（万方平台首次上网日期，不代表论文的发表时间）