Real-Time Pest Detection and Identification System Using Deep Learning
作者
Mrs. Akila M, Nithish Kumar NC, Vishnu K
出处
期刊:International Journal of Advanced Research in Science, Communication and Technology [Shivkrupa Publication's] 日期:2026-03-23卷期号:: 8-8被引量:1
标识
DOI:10.48175/ijarsct-31902
摘要
Agriculture faces significant economic losses due to pest infestations, with timely and accurate identification being a critical challenge for farmers and agronomists. Traditional pest detection relies heavily on manual inspection by experts, which is time-consuming, costly, and impractical at scale. This paper proposes an end-to-end intelligent pest detection system leveraging YOLOv8-based object detection to identify three common agricultural pests—aphids, fruit flies, and scale insects—directly from smartphone images. The trained model is served via a FastAPI backend exposed through ngrok tunneling, enabling a native Android mobile application to provide farmers with real-time pest identification and confidence scores without expert involvement. The system achieves a mean average precision (mAP@0.5) of 48.3%, a precision of 57%, and a recall of 46.9% on the test dataset, demonstrating a viable foundation for scalable precision agriculture