Scholarlink

Corn, also known as maize, is a vital crop globally, with significant economic and nutritional value. However, maize production faces major challenges owing to various diseases that cause substantial crop losses and threaten food security. Advancements in technology, particularly the widespread availability of smart devices offers a promising solution for mitigating these losses through automatic disease detection. This study introduces a method for recognizing corn leaf diseases using a DCNN. The proposed model incorporates SHAP (Shapley Additive Explanations) for enhanced explainability, allowing for better understanding and interpretation of the diagnostic process.The experimental results demonstrate that the custom Deep CNN model achieves an accuracy of 97.53%, highlighting its potential for effective and reliable disease detection in maize crops. This approach is designed to assist farmers and agricultural stakeholders in the early detection and management of diseases, thereby fostering enhanced crop productivity and contributing to the overall improvement of food security at a broader scale.

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 of Corn Leaf Diseases Using Deep CNN with SHAP-Based Explainability for Enhanced Agriculture

  • 章节

  • 发表日期：

    2025年

查看全文
  • Maharin Afroj,
  • Israt Jahan,
  • Md. Raisul Islam,
  • Md. Polash Islam

摘要

Corn, also known as maize, is a vital crop globally, with significant economic and nutritional value. However, maize production faces major challenges owing to various diseases that cause substantial crop losses and threaten food security. Advancements in technology, particularly the widespread availability of smart devices offers a promising solution for mitigating these losses through automatic disease detection. This study introduces a method for recognizing corn leaf diseases using a DCNN. The proposed model incorporates SHAP (Shapley Additive Explanations) for enhanced explainability, allowing for better understanding and interpretation of the diagnostic process.The experimental results demonstrate that the custom Deep CNN model achieves an accuracy of 97.53%, highlighting its potential for effective and reliable disease detection in maize crops. This approach is designed to assist farmers and agricultural stakeholders in the early detection and management of diseases, thereby fostering enhanced crop productivity and contributing to the overall improvement of food security at a broader scale.