Towards eco-friendly apple farming: Real-time codling moth monitoring using improved YOLOv10 and IoT integration

Abstract

Pest-related crop losses pose a critical threat to food security and sustainable agriculture, especially in apple orchards where the codling moth (Cydia pomonella) is a major concern. This study introduces an advanced pest monitoring system that integrates an improved YOLOv10-m deep learning model with Internet of Things (IoT) technology, designed specifically for real-time detection of codling moths. The system operates on a low-power Raspberry Pi platform, making it accessible and cost-effective for widespread field deployment. By enabling precise, geolocated, and real-time monitoring of pest populations, the system facilitates the rational and timely application of pesticides-only when and where they are truly needed. This not only enhances the effectiveness of pest control but also significantly reduces excessive chemical usage, thereby minimizing harmful residues in the environment and promoting better human health outcomes. Comparative evaluation against YOLO versions 5-12 confirms the superior balance of accuracy, confidence stability, and computational efficiency of the proposed model. Aligned with the principles of Integrated Pest Management (IPM), this approach promotes eco-friendly and health-conscious farming practices. Ultimately, the study demonstrates the potential of combining AI and IoT technologies to revolutionize pest management, contributing to a more sustainable and responsible agricultural ecosystem.

Abstract from the paper’s OpenAlex record; open copy licensed cc-by. Read at the source ↗

The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.