基于圆投影与径向投影的模板匹配算法

Template Matching Algorithm Based on Circular Projection and Radial Projection

摘要: 针对传统的基于相关性的模板匹配算法计算量大和不具有旋转不变性的问题，提出一种基于圆投影与径向投影的快速模板匹配算法。首先采用跳跃式步长和高斯金字塔策略快速获得圆投影向量，接着使用NCC相似性度量方法进行粗匹配；然后再进行精匹配，得到模板图像在待测图像中的准确位置；最后利用径向投影得到模板与待匹配图像之间旋转的角度。实验结果表明：基于圆投影与径向投影的模板匹配算法不仅可以解决旋转不变性的问题，还可以准确获得模板与待匹配图像之间的旋转角度，匹配速度也较快。

Abstract: Aiming at the problems of a large amount of computation and no rotation invariance of the traditional template matching algorithm based on correlation, a fast template matching algorithm based on circular projection and radial projection is proposed. Firstly, the circular projection vector is quickly obtained by using jump step size and Gaussian pyramid strategy, and then the NCC similarity measurement method is used for rough matching; then, the accurate position of the template image in the image to be measured is obtained by fine matching; Finally, the rotation angle between the template and the image to be matched is obtained by radial projection. The experimental results show that the template matching algorithm based on circular projection and radial projection can not only solve the problem of rotation invariance, but also accurately obtain the rotation angle between the template and the image to be matched, and the matching speed is also fast.