Scholarlink

Diseases on plant leaves reduces the production of crops since it severely damages the plant’s growth. Manual approaches for identifying the diseases on plant leaves needs expert’s opinion and it is a time consuming process. To overcome this issue, an automated system is required to detect the diseases on plant leaves. Hence, a method using MobileNetV2 architecture is trained to detect and classify the various diseases occurring on the leaves of plants. The Kaggle dataset is used for the training, validation and testing phase. The test results revealed the highest accuracy of using MobileNetV2 compared to GoogleNet, VGG-16, AlexNet and VGG-19. The best performance output of these five networks will be identified after calculating the evaluation parameters like Accuracy, Sensitivity, Specificity, Precision and ROC AUC. The lightweight model can be implemented in the mobile phone thereby reducing the burdens of farmers. The farmers can easily identify the diseases on plant and can take appropriate measures to reduce it.

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

A Comparative Analysis of Different Deep CNN Model for Disease Detection and Classification on Plant Leaves

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • P. Maheswari,
  • B. Ramasubramanian,
  • C. Maria Shivani,
  • A. Lerin,
  • R. Atchaya

摘要

Diseases on plant leaves reduces the production of crops since it severely damages the plant’s growth. Manual approaches for identifying the diseases on plant leaves needs expert’s opinion and it is a time consuming process. To overcome this issue, an automated system is required to detect the diseases on plant leaves. Hence, a method using MobileNetV2 architecture is trained to detect and classify the various diseases occurring on the leaves of plants. The Kaggle dataset is used for the training, validation and testing phase. The test results revealed the highest accuracy of using MobileNetV2 compared to GoogleNet, VGG-16, AlexNet and VGG-19. The best performance output of these five networks will be identified after calculating the evaluation parameters like Accuracy, Sensitivity, Specificity, Precision and ROC AUC. The lightweight model can be implemented in the mobile phone thereby reducing the burdens of farmers. The farmers can easily identify the diseases on plant and can take appropriate measures to reduce it.