Cotton Leaf Disease Classification by Combining Color and Texture Feature-based Approach

Abstract

The cotton crop is one of the leading crops grown in India, so in this case, loss of crop leads to yield loss. There are various reasons for crop loss, but one of them is leaf disease. Therefore, disease classification was considered in this study. Although different methodologies, such as k-means and ostus thresholds, have been introduced, the accuracy of the systems did not outperform. Different leaf diseases, such as Alternaria, bacterial blight, and grey mildew, have been identified using image-processing techniques. The early detection of these diseases is a challenge. Cotton leaf images with a complex background were extracted using a modified factorization-based active contour method. Segmented images, color, gray-level co-occurrence matrix, and local binary pattern features were extracted. Machine-learning classifiers were applied to the feature set for classification. The performance of the classifiers was compared to that of other state-of-the-art classifiers. From the experimental results, it is observed that the proposed ensemble classifier performed better than the single-classifier support vector machine, random forest, multilayer perceptron, and even the ensemble classifier support vector machine with a random forest with a multilayer perceptron. The accuracy of the proposed model is 92.29