Scholarlink

India is predominantly an agricultural country, with agriculture playing a crucial role in its economy and society by providing employment and livelihoods to a significant portion of the population. However, the sector faces substantial losses due to plant diseases caused by various pathogens, including worms, bacteria, viruses, and fungi. Predominant diseases like black rot, early blight, and late blight can lead to severe damage if not properly addressed. This research investigates the potential of several deep learning models, including ResNet50, InceptionV3, and MobileNet, and conducts a comparative analysis of their performance metrics. The proposed deep CNN model uses ResNet50 for feature extraction and more layers of convolutions are added on top of it. The findings demonstrate that it has achieved a remarkable testing accuracy of 97.05%, outperforming the other deep learning models examined in this study.

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

Plant Leaf Disease Detection and Severity Assessment Using Deep CNN Architectures

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Anita Shrotriya,
  • Khushi Soni,
  • Shruti Paliwal,
  • Sunita Singhal

摘要

India is predominantly an agricultural country, with agriculture playing a crucial role in its economy and society by providing employment and livelihoods to a significant portion of the population. However, the sector faces substantial losses due to plant diseases caused by various pathogens, including worms, bacteria, viruses, and fungi. Predominant diseases like black rot, early blight, and late blight can lead to severe damage if not properly addressed. This research investigates the potential of several deep learning models, including ResNet50, InceptionV3, and MobileNet, and conducts a comparative analysis of their performance metrics. The proposed deep CNN model uses ResNet50 for feature extraction and more layers of convolutions are added on top of it. The findings demonstrate that it has achieved a remarkable testing accuracy of 97.05%, outperforming the other deep learning models examined in this study.