Scholarlink

Nowadays, many plants are causing diseases due to climatic changes. The main plant that causes more diseases is the tomato, which is used in large quantities for cooking. This study focuses on the early detection and prediction of tomato leaf disease by using image classification and various deep learning techniques. If we detect the disease and give suitable measures for prediction, it will be very helpful for farmers to get good yields and also lead to sustainable development. The dataset is taken from Kaggle, and a few of the fields are located in Andhra Pradesh. Convolutional Neural Network (CNN), Residual Network (ResNet50), VGG16, and AlexNet are the techniques used to detect the disease with and without augmentation. The CNN model is found to be the best with an accuracy of 94 percent and also predicts the images correctly. The study suggests that expanding the dataset across other techniques can also give better results.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Early Tomato Leaf Disease Prediction Using Deep Learning Techniques

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Devapatla Bhargavi,
  • Kancharla Manikanta,
  • Mamta Arora

摘要

Nowadays, many plants are causing diseases due to climatic changes. The main plant that causes more diseases is the tomato, which is used in large quantities for cooking. This study focuses on the early detection and prediction of tomato leaf disease by using image classification and various deep learning techniques. If we detect the disease and give suitable measures for prediction, it will be very helpful for farmers to get good yields and also lead to sustainable development. The dataset is taken from Kaggle, and a few of the fields are located in Andhra Pradesh. Convolutional Neural Network (CNN), Residual Network (ResNet50), VGG16, and AlexNet are the techniques used to detect the disease with and without augmentation. The CNN model is found to be the best with an accuracy of 94 percent and also predicts the images correctly. The study suggests that expanding the dataset across other techniques can also give better results.