Enhancing Precision Agriculture Pest Control: A Yolov10-Based Deep Learning Approach For Insect Detection

Abstract

Precision Agriculture (PA) integrates modern technological solutions to enhance resource efficiency while maintaining crop productivity and quality. Despite these advancements, pest infestations continue to pose a significant threat to agricultural sustainability. Recent developments in deep learning-based object detection models, particularly the YOLO (You Only Look Once) series, have enabled real-time insect detection. However, many existing approaches, including YOLOv8-based systems, are often constrained to specific insect classes or crop environments, limiting their general applicability.

To overcome these challenges, this study proposes a more generalized and efficient insect detection framework utilizing the latest YOLOv10 architecture. The proposed system is designed to identify multiple insect categories across a wide range of crops, thereby supporting scalable and real-time pest monitoring in diverse agricultural settings. Experimental evaluation was conducted using a standard benchmark insect dataset to assess the model’s performance.

The results demonstrate that the YOLOv10-based approach outperforms earlier models, including YOLOv8, in terms of mean Average Precision (mAP) and inference speed. These improvements can be attributed to the architectural enhancements in YOLOv10, which enable better feature representation and faster processing. Overall, the proposed method offers a robust and adaptable solution for pest detection, contributing to more effective pest management strategies in precision agriculture.