Study on Aircraft Inspection Based on Improved YOLOv8
This paper proposes an aircraft detection method based on the improved YOLOv8 model,aiming to improve the accuracy and real-time performance of detection.First,this paper introduces the inverse residual attention module(iRMB)to enhance the model's ability to learn aircraft features through an improved attention mechanism.Second,the cen-tered feature pyramid(EVC)module is used to optimize the feature extraction process and enhance the model's ability to detect aircraft at different scales.In addition,an improved distance intersection and merger ratio(MDIoU)is adopted as the loss function in this paper to further enhance the localization accuracy of the model.Experimental results on the aircraft category of the publicly available dataset Caltech101 show that compared with existing aircraft detection methods,the method proposed in this paper achieves 98.2%and 98.9%in terms of detection precision and recall,respectively,and per-forms more prominently especially in complex background and multi-scale target detection scenarios.
万方数据
维普
