Scholarlink

<p>To keep farms productive and cut down on money lost, it’s important to quickly and accurately identify plant diseases. This study introduces deep learning-based framework for classifying diseases of apple leaves. The system uses a large dataset that has six different classes: healthy leaves, diseases like Apple Black Rot, Apple Cedar Rust, and Apple Scab to test different deep learning models and their combinations. With an accuracy of 96.84%, RESNET50V2 outperformed SVM (75.93%) and LR (56.54%) among the individual models. With an accuracy of 97.05%, a combination of RESNET50 and RESNET50V2 also showed the best performance, demonstrating the potential of ensemble techniques in illness classification tasks. The study tackles issues like class disparity and shows how well the suggested method works for a variety of ailments, especially for rare conditions like Apple Powdery Mildew Complex. Using real-time with IoT devices, growing datasets, and incorporating explainable AI techniques to encourage transparency are some future directions. The suggested strategy promotes efficient and sustainable agricultural practices by providing a solid basis for improving disease management in apple agriculture.</p>

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

Deep learning based classification of apple leaf diseases using image analysis

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月1日

查看全文
  • Hirenkumar Kukadiya,
  • Divyakant Meva,
  • Awniya Kumar

摘要

To keep farms productive and cut down on money lost, it’s important to quickly and accurately identify plant diseases. This study introduces deep learning-based framework for classifying diseases of apple leaves. The system uses a large dataset that has six different classes: healthy leaves, diseases like Apple Black Rot, Apple Cedar Rust, and Apple Scab to test different deep learning models and their combinations. With an accuracy of 96.84%, RESNET50V2 outperformed SVM (75.93%) and LR (56.54%) among the individual models. With an accuracy of 97.05%, a combination of RESNET50 and RESNET50V2 also showed the best performance, demonstrating the potential of ensemble techniques in illness classification tasks. The study tackles issues like class disparity and shows how well the suggested method works for a variety of ailments, especially for rare conditions like Apple Powdery Mildew Complex. Using real-time with IoT devices, growing datasets, and incorporating explainable AI techniques to encourage transparency are some future directions. The suggested strategy promotes efficient and sustainable agricultural practices by providing a solid basis for improving disease management in apple agriculture.