Comparative Study of Cotton Leaf Disease Classification

Abstract

50% of the Indian population depends on the cotton crops for their business. The recognition of cotton disease is very important because it decreases the quality and productivity of the cotton. It starts with detecting the diseases by collecting the images of the diseased leaf. In this paper the major cotton leaf disease are alternaria, bacterial blight, cercospra, and Myrothecium leaf spot. Few of the common basic steps to detect the cotton disease have been used in this paper such as image acquisition, pre-processing, feature extraction and classification. The methods used for the segmentation are k-means clustering and Otsu’s threshold. The ANN and SVM are the two methods which we have used most for classifying the type of the cotton disease. The ANN and SVM provide the accuracy of 80-90%.