Scholarlink

Plant diseases represent a significant challenge in agriculture, significantly impacting crop yield. Addressing this issue is imperative to ensure agricultural productivity. Convolutional Neural Networks (CNNs) have emerged as a promising solution for image-based disease detection. Detecting diseases in their early stages using CNN models can potentially enhance crop yield for farmers. Given the multi-class nature of most plant diseases, we conducted a comprehensive comparative study of CNNs to assess their effectiveness. Our study compared three CNN models: AlexNet, VGG-19, and ResNet18. ResNet18 demonstrated superior performance, achieving a validation accuracy of 95.68%. Furthermore, we developed a web application to enable farmers to utilize the trained model effectively.

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

Optimizing Crop Health: Evaluating CNN Performance in Plant Disease Detection

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Parimi Charan Tejas,
  • Marnala Murali Sriram Ganapathi,
  • Parepalli Mahesh Babu,
  • Kosuru Surya Sai,
  • Godi Amulya,
  • Sridevi Bonthu

摘要

Plant diseases represent a significant challenge in agriculture, significantly impacting crop yield. Addressing this issue is imperative to ensure agricultural productivity. Convolutional Neural Networks (CNNs) have emerged as a promising solution for image-based disease detection. Detecting diseases in their early stages using CNN models can potentially enhance crop yield for farmers. Given the multi-class nature of most plant diseases, we conducted a comprehensive comparative study of CNNs to assess their effectiveness. Our study compared three CNN models: AlexNet, VGG-19, and ResNet18. ResNet18 demonstrated superior performance, achieving a validation accuracy of 95.68%. Furthermore, we developed a web application to enable farmers to utilize the trained model effectively.