面向航线检修的飞机轮胎缺陷检测算法OA

In allusion to the problem that the complex and diverse defect patterns of aircraft tires in aviation maintenance scenarios make it difficult for existing detection methods to strike a balance between accuracy and model size,an improved YOLOv8-based aircraft tire defect detection algorithm,MAME-YOLO,is proposed.A MESSA-C2f module is designed to enhance multi-scale feature extraction and improve the recognition of irregular defects.The ADown module is introduced to optimize the downsampling process,enhancing feature representation while effectively reducing model parameters.The MASAFPN neck network is constructed to achieve adaptive multi-scale feature fusion of the objects,strengthening the detection of small-scale defects.The EDTAH detection head is designed to optimize the decoupling mechanism,improving discrimination in complex backgrounds and further compressing model parameters.In order to validate the method,experiments are conducted on a self-built aircraft tire defect dataset.The results show that,in comparison with the baseline model,MAME-YOLO can realize a 5.3%increase in mAP@0.5 and a 23.6%reduction in model parameters.This method can provide a feasible technical solution for fast and accurate detection of aircraft tire damage in line maintenance,with promising prospects for engineering applications.