Scholarlink

Tomato leaf diseases can drastically influence crop yield and quality, resulting in considerable economic markdown for farmers. Timely diagnose and precise classification of these pathologies are critical for efficient administration. This paper explores the classification of tomato leaf diseases using four deep learning models: CNN, ResNet-50, InceptionV3, and EfficientNetB0. Utilizing images from the Plant Village database, we assess the model’s performance with two optimizers, Adam and Nadam. EfficientNetB0 stands out, achieving the highest accuracy of 99%, demonstrating its superior capability in handling the classification task with low computational cost. This study underscores the importance of advanced deep learning techniques and optimizers in enhancing disease detection accuracy. Furthermore, we examined the role of transfer learning in improving model accuracy and training efficiency. This study provides valuable understanding of effectiveness of various CNN architectures and optimizers, providing practical solutions for the timely recognition of diseases in tomato leaves. Ultimately, these findings support farmers in better crop management and help mitigate economic losses.

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

Detection and Classification of Tomato Crop Disease Using Deep Learning Models with Varied Optimization Techniques

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Abhishek Rana,
  • Neelam Goel

摘要

Tomato leaf diseases can drastically influence crop yield and quality, resulting in considerable economic markdown for farmers. Timely diagnose and precise classification of these pathologies are critical for efficient administration. This paper explores the classification of tomato leaf diseases using four deep learning models: CNN, ResNet-50, InceptionV3, and EfficientNetB0. Utilizing images from the Plant Village database, we assess the model’s performance with two optimizers, Adam and Nadam. EfficientNetB0 stands out, achieving the highest accuracy of 99%, demonstrating its superior capability in handling the classification task with low computational cost. This study underscores the importance of advanced deep learning techniques and optimizers in enhancing disease detection accuracy. Furthermore, we examined the role of transfer learning in improving model accuracy and training efficiency. This study provides valuable understanding of effectiveness of various CNN architectures and optimizers, providing practical solutions for the timely recognition of diseases in tomato leaves. Ultimately, these findings support farmers in better crop management and help mitigate economic losses.