Study And Application Of Image Matching Based On Election Campaign Optimization Algorithm

The technology of image matching plays an important role in the field of image processing, and accounted for great proportion in the application of the machine vision. The image matching is based on the real-time image and the reference image to choose some features, search strategy, operational similarity to determine the best space matching corresponding points. The evaluation criteria of image matching include matching probability, matching velocity and matching accuracy, which were mainly influenced by feature space, similar inertia, search space and search strategy.Image matching algorithm is chiefly divided into gray correlation image matching algorithm and feature points based image matching. Because of the search strategy of pixel by pixel traversal, the amount of calculation of them are relatively large and matching speed are relatively slow, which lead the gray correlation image matching methods be not applicable. In this paper, the problem of image matching will be transformed into a global optimization problem. Then find the global optimal solution through Election Campaign Optimization Algorithm, which will accelerate the matching speed and optimum the matching performance.The Election Campaign Optimization Algorithm is swam intelligence heuristic global optimization algorithm with the characteristics of not restricted by initial value and information sharing. The mechanism of parallel search in global scope can not only accelerate the search speed, but also avoid fall into the local optimal trap. The Election Campaign Optimization Algorithm can directly apply to the gray image matching, but image matching based on feature points has to be corresponding changed. Improved process will match, in search space of source image, the feature points with the feature points found in temp image, that the Election Campaign Optimization Algorithm will play a vital role in this process and then improve the matching performance.According to the theory above, this paper has finished related work. Firstly, the Election Campaign Optimization Algorithm based on VC++has been realized, which parameterize the input mode of the objective function and facilitate the software. Secondly, the gray image matching based Election Campaign Optimization Algorithm and feature points image matching based Election Campaign Optimization Algorithm have been realized. And this paper also complete the contrast experiment between traditional matching algorithm and improved algorithm, which verify Election Campaign Optimization Algorithm has the advantages that it can improve the matching speed and optimize the matching performance. Finally, the Election Campaign Optimization Algorithm has been applied to the machine vision system that used to detect the defects of mark, which make full use of the advantages of image matching based on Election Campaign Optimization Algorithm.