Research Of Monitoring Corn Diseases Method Based On Machine Vision

China is an agricultural country,and the agriculture is one of the main incomes for economy.Corn is affected by pests and diseases from time to time,and it restricted crop yields.It is effective to reduce lose by recognizing crop diseases immediately.Corn is one of the most crops in China.In recent years,the corn yields are declined,so computer vision technology can be used to identify different kind of common corn diseases accurately and aim to control corn diseases.So the paper will be surrounded by using the Computer vision of image processing technology to recognize different kind of corn diseases and simulation experiment.First of all,the four kinds of corn diseases(grey speck disease,NCLB disease,Bipolaris maydis,rust disease)are selected as research object by camera,and the non-relevant variables(shadow,illumination etc.)were eliminated.Based on this,the processed images were segmented by a color clustering algorithm and the edge of disease areas and the character of color are saved well.In addition,the method of color,texture and shape characters were extracted for corn images.The parameters of the first moment,the second moment,the third moment were gotten by color features;The parameters of entropy contrast,correlation,entropy,homogeneity,energy are obtained by texture features;The parameters of the area,circumference,rectangle degree,elongation,roundness are acquired by shape features,so the difference of four corn diseases can be classified accurately.At last,the method of support vector machine is used to identify corn diseases;Based on SVM,The data is classified and database is built to acquire the last recognition results.The result shows that it existed several errors by single color feature,texture feature,shape feature,however the method of multiple features can get a desired outcome;Based on this,the visual interface of corn diseases were designed by using vs2013 and the interaction was also good,so it provided reference for the identification of maize leaf disease.The image processing technology is used to identify four types of corn common diseases,the lesion areas are segmented by the method of clustering algorithm and combined eigenvector of image texture to make statistical and classify,and it has practical value.