典型文献
Visual inspection of aircraft skin:Automated pixel-level defect detection by instance segmentation
文献摘要：
Skin defect inspection is one of the most significant tasks in the conventional process of aircraft inspection.This paper proposes a vision-based method of pixel-level defect detection,which is based on the Mask Scoring R-CNN.First,an attention mechanism and a feature fusion module are introduced,to improve feature representation.Second,a new classifier head—consisting of four convolutional layers and a fully connected layer—is proposed,to reduce the influence of informa-tion around the area of the defect.Third,to evaluate the proposed method,a dataset of aircraft skin defects was constructed,containing 276 images with a resolution of 960×720 pixels.Experimental results show that the proposed classifier head improves the detection and segmentation accuracy,for aircraft skin defect inspection,more effectively than the attention mechanism and feature fusion module.Compared with the Mask R-CNN and Mask Scoring R-CNN,the proposed method increased the segmentation precision by approximately 21％and 19.59％,respectively.These results demonstrate that the proposed method performs favorably against the other two methods of pixel-level aircraft skin defect detection.
文献关键词：
中图分类号：
作者姓名：
Meng DING;Boer WU;Juan XU;Abdul Nasser KASULE;Hongfu ZUO
作者机构：
College of Civil Aviation,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China;Key Laboratory of Civil Aircraft Health Monitoring and Intelligent Maintenance of CAAC,Nanjing 210016,China
文献出处：
引用格式：
[1]Meng DING;Boer WU;Juan XU;Abdul Nasser KASULE;Hongfu ZUO-.Visual inspection of aircraft skin:Automated pixel-level defect detection by instance segmentation)[J].中国航空学报（英文版）,2022(10):254-264
A类：
B类：
Visual,inspection,aircraft,skin,Automated,level,detection,by,instance,segmentation,Skin,one,most,significant,tasks,conventional,process,This,paper,proposes,vision,which,Mask,Scoring,First,attention,mechanism,feature,fusion,module,introduced,representation,Second,new,classifier,head,consisting,four,convolutional,layers,fully,connected,proposed,reduce,influence,informa,around,area,Third,evaluate,dataset,defects,was,constructed,containing,images,resolution,pixels,Experimental,results,show,that,improves,accuracy,more,effectively,than,Compared,increased,precision,approximately,respectively,These,demonstrate,performs,favorably,against,other,two,methods
AB值：
0.511183
相似文献
机标中图分类号，由域田数据科技根据网络公开资料自动分析生成，仅供学习研究参考。
