Abstract: Pine Wilt Disease(PWD)is a devastating disease that severely harms pine forest ecosystems.It spreads rapidly and is difficult to control,while traditional manual monitoring methods are inefficient and costly.To address this,an improved lightweight object detection network,M-YOLOv8n-p2,is proposed based on drone aerial image data.By introducing a multi-kernel feature fusion perception mechanism,the model's ability to distinguish features of dead pine trees is effectively enhanced.Compared with existing YOLO improvement methods,while maintaining lightweight,it significantly enhances the extraction and fusion efficiency of multi-scale features,effectively solving the problem of missed and false detections of small targets in complex backgrounds.The proposed network has higher accuracy and robustness in the detection task of dead trees caused by Pine Wilt Disease,enabling efficient and accurate identification of diseased trees.It can provide reliable technical support for intelligent monitoring and prevention and control of forestry pests and diseases.