Research On Dairy Cow Insurance System Based On Pattern Recognition

This discourse is about the cow back edges matching technology in the area of agricultural insurance. The background is the practical application in cow insures and cow compensation.In view of the fact that cow back design particularity, this discourse proposed a precise matching algorithm between two images based on epipolar constraint and edge points detecting. in the case of two specified images matching, after two images epipolar constraint variant, matching them through cows' back edges. The epipolar geometry and edge matching technology are mainly used in the algorithm.In case of the numbers lost, this discourse also proposed a dairy cow search algorithm, by extract huge edges from cow back edges, to search in a wide range of cows images. After filter, the candidates would be matches in the phase correlation function, to improve the retrieval speed and efficiency.The feasibility and effectiveness of the two algorithms have been fully discussed in this discourse, and a complete cow back pattern matching system has been set up. It been sure that the realization of algorithm would provide solid theoretical basis and practical value in dairy cow identification and underwriting of agriculture insurance company.