Scholarlink

Codling moth detection in apple orchards relies on labor-intensive manual inspection methods, leading to delayed responses and significant crop losses. Existing automated pest monitoring systems lack integration between modern deep learning and practical IoT deployment for real-time agricultural applications. This paper presents an IoT-integrated smart trap system combining deep learning detection with Raspberry Pi edge computing and cloud based management for automatic codling moth monitoring. The system uses camera equipped traps for continuous surveillance with automated cloud dashboard alerts when detection thresholds are exceeded. Comparative evaluation of YOLO architectures (v8, v10, v11, v12) under identical conditions identified YOLOv10-m as optimal, achieving 100% precision with only 16.48 million parameters for efficient edge deployment. The system transforms reactive pest control into proactive, data-driven management, reducing chemical dependency while providing real-time actionable insights. This work contributes an end-to-end solution bridging advanced computer vision with practical agricultural implementation, demonstrating effective deep learning deployment on IoT edge devices for precision agriculture.

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

YOLOv10-Based Smart Trap for Codling Moth Detection with IoT Alerts

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Asma Ait Ba,
  • Mohamed Zarboubi,
  • Fatima Hmache,
  • Mariam Chahba,
  • Samira Chabaa,
  • Azzedine Dliou,
  • Abdelouhab Zeroual,
  • Abdelaaziz Bellout,
  • Amine Saddik,
  • Rachid Bouharroud

摘要

Codling moth detection in apple orchards relies on labor-intensive manual inspection methods, leading to delayed responses and significant crop losses. Existing automated pest monitoring systems lack integration between modern deep learning and practical IoT deployment for real-time agricultural applications. This paper presents an IoT-integrated smart trap system combining deep learning detection with Raspberry Pi edge computing and cloud based management for automatic codling moth monitoring. The system uses camera equipped traps for continuous surveillance with automated cloud dashboard alerts when detection thresholds are exceeded. Comparative evaluation of YOLO architectures (v8, v10, v11, v12) under identical conditions identified YOLOv10-m as optimal, achieving 100% precision with only 16.48 million parameters for efficient edge deployment. The system transforms reactive pest control into proactive, data-driven management, reducing chemical dependency while providing real-time actionable insights. This work contributes an end-to-end solution bridging advanced computer vision with practical agricultural implementation, demonstrating effective deep learning deployment on IoT edge devices for precision agriculture.