REAL-TIME INSECT DETECTION FOR PRECISION AGRICULTURE AND ECOLOGICAL MONITORING

Abstract

The remarkable achievement of computer vision technology is the automatic detection of insects in the field of agriculture and ecological monitoring. Effective and quick identification of species is helpful for sustainable farming. It helps in the use of pesticides as well as in mitigating crop failures. Traditional images processing techniques performs poorly in different environments. Deep learning (DL) in Artificial Intelligence (AI) has automated many processes by increasing the accuracy and robustness of the system. This study focuses on the use of the real-time insect detection ability of one of the most famous detectors, YOLOv11, under the open field scenario. The model can perform well-accurate detections with improved feature pyramid structures. The model has achieved an average precision (mAP@.50) of 98.2%. This result shows that Yolov11 is highly accurate and optimal for the detection of insects in agriculture, ecological monitoring missions, and integrated pest management.

Keywords : Insect detection, Computer vision, Deep learning, Pest management, Real-time monitoring, Precision agriculture.