SUGARCANE DISEASE IDENTIFICATION USING CONVOLUTION NEURAL NETWORK

Abstract

In the contemporary era, the upsurge of plant ailments stands out as a significant factor impacting diminished agricultural yield and escalating losses among farmers. Thus, it becomes paramount to utilize a method capable of furnishing prompt and accurate outcomes. The recent proliferation of deep learning has emerged as pivotal in effectively addressing both traditional and unconventional hurdles. The Convolutional Neural Network (CNN) has surfaced as an advanced strategy for cutting-edge identification and detection. In confronting the challenge of plant diseases, we have curated an exhaustive dataset encompassing 37 distinct plant disease varieties corresponding to 5 seedling varieties (238, 223, 268, 218, 214) for the training and validation of our model. Our approach entails the utilization of Resnet (Residual Neural Network), a specialized CNN architecture. We capture images of afflicted plant foliage and leverage CNN-driven classification for disease identification. Our model showcases superior precision in contrast to various previously employed techniques for disease detection.