Scholarlink

To boost agricultural productivity we must be efficient in disease management of crops. The paper aims to aid Indian farmers in quickly detecting and curing agricultural diseases by applying state-of-the-art Deep Learning algorithms for classifying diseases in chili leaves. In this study, the authors evaluated the performance of ResNet, MobileNet, and VGG16 architectures proposed detecting the four categories of a leaf, Powdery Mildew, Bacterial Spot, Leaf Curl, and Healthy Leaves on thermal spectrogram-based images. To ensure the model trains correctly, the dataset containing images of different diseases is put through preprocessing [2] steps like scaling and standardization. Every model makes a serious pain in tuning the optimal hyperparameters in cause to improve the accuracy of prediction. The models are benchmarked with various evaluation metrics like F1-score, recall, accuracy, and precision. Candidatus Liberibacter asiaticus and Xylella fastidiosa. The three models selected were able to correctly identify these two diseases of chili leaves. This work contributes to the development of precision agriculture since optimizing the crop yield and quality, indicating a scalable disease classification and enables early treatments.

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

Chili Leaves Disease Detection Using Thermal Images and Deep Learning Classification

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • S. N. V. V. S. Gowtham Tadavarthy,
  • Changala Jayesh Reddy Eddula,
  • Ravi Chandu Bollepalli,
  • Anindh Varma Datla,
  • S. Sarath

摘要

To boost agricultural productivity we must be efficient in disease management of crops. The paper aims to aid Indian farmers in quickly detecting and curing agricultural diseases by applying state-of-the-art Deep Learning algorithms for classifying diseases in chili leaves. In this study, the authors evaluated the performance of ResNet, MobileNet, and VGG16 architectures proposed detecting the four categories of a leaf, Powdery Mildew, Bacterial Spot, Leaf Curl, and Healthy Leaves on thermal spectrogram-based images. To ensure the model trains correctly, the dataset containing images of different diseases is put through preprocessing [2] steps like scaling and standardization. Every model makes a serious pain in tuning the optimal hyperparameters in cause to improve the accuracy of prediction. The models are benchmarked with various evaluation metrics like F1-score, recall, accuracy, and precision. Candidatus Liberibacter asiaticus and Xylella fastidiosa. The three models selected were able to correctly identify these two diseases of chili leaves. This work contributes to the development of precision agriculture since optimizing the crop yield and quality, indicating a scalable disease classification and enables early treatments.