Groundnut leaf disease identification using image processing

Abstract

In Indian agriculture, the most common oilseed crop is groundnut. Disease attacks are one of the most significant reasons that lead to a low agricultural commodity. When plants are infected with different diseases through their leaves, it affects farm productivity and results in a profit loss. Plant diseases such as fungi, bacteria, and viruses reduce crop yields. Cercospora (an early leaf spot), late leaf spot, rust, and Alternaria leaf spot are fungal diseases that damage groundnut crops. The earlier identification of plant diseases plays a vital role in overcoming low-quality product yield. Convolutional Neural Networks (CNN) model aids in detecting the groundnut leaf disease, and the k-means clustering method assist the segmentation. Images of the infected groundnut leaf disease are identified by extracting the features from the segmented image after the pre-processing stage is performed. CNN classifier training and testing demonstrate accurate groundnut disease classification results in 93 % accuracy.