Analysis Of The Disease Identification System Of Pear Tree Leaves Based On Image Analysis

The pear production scale is very large in our country,but the quality and the yield are seriously affected by the pear diseases.According to statistics,there are a great variety of pear diseases,more than 80 kinds.Among all of them,the harms caused by the black spot disease,the brown spot and the Ventura are relatively large.The diseases resulted in large-scale reduction of pear production.It not only has a direct impact on economic income of farmers,but also limits the expansion of cultivated areas.With the software of MATLAB,combing the methods of disease identification and the technology of image analysis and recognition together could identity the type of pear disease in leaves.The main contents and progress of the study are as follows:1 In the aspect of disease image enhancement,image gray transformation,histogram equalization and gray level conversion are the first steps in image processing.Its effects of digital image conversion are quite different from which of PS and the like.The purposes of histogram equalization are to make the image gray distribution uniform and get the clearer images.In lesion segmentation of the image,the method of thresholding segmentation can directly segment the colorful images and the segmentation effect is great.2 After the completion of image processing,the mage feature extraction can be done from the following three aspects: texture features,shape features and color features.The extracted datas are used as the database to construct the disease recognition model.At last,with the built data model,inputting images to be detected,the correct rate of diseases recognition of pear leaves could up to 91.7%.3 The entire system is implemented in MATLAB which is used to program.The farmers could identify the type of pear diseases by themselves just via taking pictures of disease part of the pear leaves with their mobile phone.The system provides a convenient and economical service for farmers,increasing the production and improving the yield of pear fruit quality,reducing the environmental pollution caused by pesticide abuse.4 Develop pear leaf disease identification web diagnostic system with computertechnology and network information technology.The system uses the browser / server as the basic network framework,taking the Zend Studio,Dreamweaver and other as code editor,choosing PHP programming language to develop,to provide users with a convenient real-time INTERNET service platform.The diversification of network information technology and the advantages of B/S architecture give the system scalability,friendly and dynamic,rich interactive man-machine interface between system and users,and improve the convenience of system maintenance and updating.