Scholarlink

Fruits are crucial for sustaining living species due to their nutritional value and role in ecosystem maintenance. A thorough inspection is necessary to ensure their quality, especially regarding leaf diseases. This study presents a model for detecting and classifying diseases in fruit leaves, utilizing five transfer learning models and a CNN (Convolutional Neural Network) model. The images underwent preprocessing before being fed into the models, including MobileNetV2, InceptionV3, ResNet50V2, Xception, VGG16, and a CNN model. MobileNetV2 and VGG16 exhibited the highest accuracy. Consequently, an ensemble model based on MobileNetV2 and VGG16 was proposed. The suggested model demonstrates significant accuracy in detecting fruit leaf diseases.

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 Detection and Classification of Fruit Leaf Diseases Using an Ensemble Approach Based on Transfer Learning

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Mariam Chowdhury,
  • Afrida Israt Jahan,
  • Saima Murtuza,
  • Tajbia Karim,
  • Md Jaber Al Nahian

摘要

Fruits are crucial for sustaining living species due to their nutritional value and role in ecosystem maintenance. A thorough inspection is necessary to ensure their quality, especially regarding leaf diseases. This study presents a model for detecting and classifying diseases in fruit leaves, utilizing five transfer learning models and a CNN (Convolutional Neural Network) model. The images underwent preprocessing before being fed into the models, including MobileNetV2, InceptionV3, ResNet50V2, Xception, VGG16, and a CNN model. MobileNetV2 and VGG16 exhibited the highest accuracy. Consequently, an ensemble model based on MobileNetV2 and VGG16 was proposed. The suggested model demonstrates significant accuracy in detecting fruit leaf diseases.