Scholarlink

Agriculture plays a crucial role in our country, and tomatoes are a popular crop globally. It is used in various forms in different cuisines worldwide. India is the second largest producer of tomatoes. However, the quality and quantity of tomato crops often suffer due to various diseases. To address this issue, use the plantVillage dataset from Kaggle, which includes nine classes of tomato leaf disease images and one class of healthy tomato leaf images. The aim is to create an automated model to help farmers detect diseases early and intervene promptly, reducing crop losses. The process involves preprocessing the input images, segmenting the targeted area from the original images and further processing the images using relevant features to effectively predict and classify the diseases. The convolutional neural network (CNN) model yielded a 92.84% accuracy, which is better than the Inception v3, VGG16, and ResNet50 models considered in the work.

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

Detection of Tomato Leaf Disease Using Convolutional Neural Network

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Manohar Madgi,
  • Vivek Verma,
  • Prabhat K. Rai,
  • Tsewang Norbu,
  • Vrashabh Kambayal

摘要

Agriculture plays a crucial role in our country, and tomatoes are a popular crop globally. It is used in various forms in different cuisines worldwide. India is the second largest producer of tomatoes. However, the quality and quantity of tomato crops often suffer due to various diseases. To address this issue, use the plantVillage dataset from Kaggle, which includes nine classes of tomato leaf disease images and one class of healthy tomato leaf images. The aim is to create an automated model to help farmers detect diseases early and intervene promptly, reducing crop losses. The process involves preprocessing the input images, segmenting the targeted area from the original images and further processing the images using relevant features to effectively predict and classify the diseases. The convolutional neural network (CNN) model yielded a 92.84% accuracy, which is better than the Inception v3, VGG16, and ResNet50 models considered in the work.