Image Recognition of Maize Diseases Based on Fuzzy Clustering and Support Vector Machine Algorithm

摘要

As to the fact that recognition rate of maize diseases is not high enough, this paper aims to find a new method used for recognize maize disease in ways of fuzzy cluster and vector machine based on machine vision technology. Enhance images with median filter method, and divide disease with the method of fuzzy c-means clustering algorithms, then extract color, shape, and texture feature. Finally, recognize disease with SVM recognition method, and the recognition accuracy rate is above 95%. Median filter algorithm could be used to smooth disease images of maize effectively. Meanwhile segmentation algorithm of fuzzy cluster could divide disease images of maize accurately, especially it is better that make use of SVM linear kernel function identifying these images as classified function.