Computer Vision based Plant Leaf Disease Recognition using Deep Learning

Abstract

Computer vision-based applications play a vital role in the era of computer science and engineering. Now-a-days peoples are facing different problems in agricultural fields to improve their cultivation. So, a better approach is proposed for plant leaf disease recognition using deep learning techniques for agricultural improvement. This research is very much helpful for the farmers to identify the leaf diseases of a plant. This proposed system has three subsections. One is feature extraction, second is trained networking generation and the third one is classification. This system first takes an image as the input and extracts the features from the image using K-means clustering. Secondly, it generates a trained network using Convolutional Neural Networks (CNNs). Then compare the original leaf image with the generated trained database in the classification section and recognition of the disease of the plant. Different techniques are used in this system for properly recognized the diseases. After analyzed the 3000 trained images, three types of leaf diseases are properly recognized by this system, which are Cercospora Leaf Spot, Mosaic virus, and Alternaria Leaf Spot. The overall accuracy of this system is very good and which is up to 95.26%.