Rotated Image Matching Based On Election Campaign Algorithm

Image matching is always a very important research topic in image processing technology, because of the different angles during the image acquisition process, the image in space there is the phenomenon such as rotation and translation, making the match more difficult. To solve this problem, commonly used method is to rotate 360 degrees with a template for method can obtain accurate matching results, it’s very slow and difficult to meet the necessary of real-time match. So looking for a fast, high accuracy, good anti-rotation matching method has become the focus of the image matching technology research.Circular isotropic, namely rotational invariance, this feature provides a good solution idea for the anti-rotation of image matching-circular projection. To solve the problem of image matching speed, we need to start from the search strategy, campaign optimization algorithm is a heuristic algorithm established on simulating human candidates campaign in pursuit of the highest approval rating motive. It uses the heuristics and traversal search strategy. At the same time it has the parallelism that can ensure search results with global optimality. Therefore, this paper presents the rotating image matching method with a circular projection to match model, simplified to campaign algorithm as the search strategy.First, we introduce the basic concept, mathematical description and classification of image matching techniques. On this basis, we deeply analyzed the four key elements of image matching and the performance evaluation of image matching.Then, we introduce the basic principles, basic characteristics of election algorithms. On this basis, we introduce the campaign calculation process of the algorithm in detail and deeply analyzed the influence factors of campaign to calculate.Finally, highlighting the rotating image matching algorithm based election campaign algorithm. First, the rotating image matching algorithmbased election campaign algorithm is divided into two matching step: matching and position matching angle; then, the four key elements of image matching from the start, the two-step matching the design of the four key elements; secondly, detailed design calculation algorithm based on a rotating campaign image matching algorithm; and again, in order to make the algorithm can be directly applied to production practice, implementation of the algorithm in VC++; and finally, matching experiments and comparative analysis of experimental results verify that the proposed algorithm is have practical value, experimental parameters for efficient application of this algorithm also provides some help.