Research on Template Matching Algorithm Based on an Improved Continuous Circular Projection

Rotation invariance is one of the important research contents of template matching algorithm. This paper mainly discusses the matching problem of target image under the condition of rotation, and proposes a template matching method based on continuous circular projection transformation. In this method, the continuous circular projection transformation model is used as the feature mapping method, and the feature vector with rotation invariance is obtained by combining the gray feature extraction method of the local gradient extremum, and the template matching is completed based on the vector. The experimental results show that the matching rate of the algorithm is 95.8% in the original data set; the occlusion matching rate of the algorithm is 98.3%_in the rotated image set and_the matching rate is 85.4% in the interference image set.