Leaf Disease Detection and Prevention Using Deep Learning

EasyChair Preprint 7621

Abstract

Since time unknown, agriculture is the backbone of Indian economy. Many plants are eaten alive by plant diseases. These plant diseases should be detected in early stages itself which helps in preventing crop damage and stops advancement of disease. Generally, plant diseases can be easily found by their visible appearance on leaves. Most Plants like groundnut, sugarcane, po­tato, orange exhibit these disease patterns on their leaves. Farmers observe these patterns and conclude to the disease caused, which is a traditional method of identifying.  The patterns that appear on leaves are taken as base to predict the related disease using some Deep Learning algorithms. In this work, Convolutional Neural Network (CNN) model is used to predict plant diseases which is highly suitable for image-based classification. This CNN model gives quicker and more proper predictions when compared to manual observation. In this work, pre trained models like Support Vector Machine, Artificial Neural Network and ResNet50 and CNN model are disciplined using the dataset. CNN model achieves more accuracy when compared to other algorithms.

Keyphrases: Artificial Neural Network, Convolutional Neural Networks, ResNet50., Support Vector Machine

@booklet{EasyChair:7621,
  author    = {Sampath Tamilselvan and Paladugu Thanuja and B. Geethavani and Eadadhala Theja and Shaik Afrin and Shaik Youneesh Basha},
  title     = {Leaf Disease Detection and Prevention Using Deep Learning},
  howpublished = {EasyChair Preprint 7621},
  year      = {EasyChair, 2022}}