Early-Stage Plant Disease Prediction Using Multilayer Convolutional Neural Network Framework

Abstract

Crop diseases are a major risk to food security, but its identification is difficult due to lack of infrastructure. Detecting and Identifying disease from the plants images is one of the interesting research areas in computer and agriculture field. The main objective in this paper is to help farmers to identify the disease and to prevent the plants in early stage. Plant disease recognition model, based on leaf image classification, developed using convolutional neural network. The developed model trained from scratch to identify the disease and it is able to recognize 10 different types of plant diseases. The reason for choosing CNN is CNN Classifier algorithm mainly extracts more features from the image datasets rather than other classifier algorithms. The trained model achieves an accuracy of 97%.