The Auto Recognition And Analysis Of Plant Black Rot Spot Based On Computer Vision

In this dissertation, we research the technology of actualize auto analysis of black rot spot. In the research, we use computer vision for the important measure, at the same time, we adopt other measures include digital image processing, artificial neural network and so on. In this paper, it include as follow:1) To design and extract the geometry character of black rot images which have been pretreated.2) To extract and analysis color characters of rot using methods of color moment and red-green pixels characters.3) To extract texture character of rot using method of two-dimension disperse wavelet transform.4) In practical application, the geometry information, color information and texture information of spot are integrated. We classify these spots by harmed degree, at last we compared this method with traditional method and got conclusion: this new method overcome the infection of personal factor and improve the decision precision.The research provide important theory basis and application technology for exploit auto-analysis of plant disease system, it also has important signification for reducing the distance of automatic manage between our country and developed countries, improving the application of computer vision technology, artificial neural network and expert system in agriculture engineering field in our country.