Railway Subgrade Defect Automatic Recognition Method Based on Improved Faster R-CNN
Railway subgrade defect is the serious threat to train safety. Vehicle-borne GPR method has become the main railway subgrade detection technology with its advantages of rapidness and nondestructiveness. However, due to the large amount of detection data and the variety in defect shape and size, defe...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Hindawi Limited
2018-01-01
|
Series: | Scientific Programming |
Online Access: | http://dx.doi.org/10.1155/2018/4832972 |
id |
doaj-9a65769585da4b828a7794d62a0bcc00 |
---|---|
record_format |
Article |
spelling |
doaj-9a65769585da4b828a7794d62a0bcc002021-07-02T07:30:06ZengHindawi LimitedScientific Programming1058-92441875-919X2018-01-01201810.1155/2018/48329724832972Railway Subgrade Defect Automatic Recognition Method Based on Improved Faster R-CNNXinjun Xu0Yang Lei1Feng Yang2School of Mechanical Electronic & Information Engineering, China University of Mining & Technology, Beijing 100083, ChinaInfrastructure Inspection Research Institute, China Academy of Railway Sciences, Beijing 100081, ChinaSchool of Mechanical Electronic & Information Engineering, China University of Mining & Technology, Beijing 100083, ChinaRailway subgrade defect is the serious threat to train safety. Vehicle-borne GPR method has become the main railway subgrade detection technology with its advantages of rapidness and nondestructiveness. However, due to the large amount of detection data and the variety in defect shape and size, defect recognition is a challenging task. In this work, the method based on deep learning is proposed to recognize defects from the ground penetrating radar (GPR) profile of subgrade detection data. Based on the Faster R-CNN framework, the improvement strategies of feature cascade, adversarial spatial dropout network (ASDN), Soft-NMS, and data augmentation have been integrated to improve recognition accuracy, according to the characteristics of subgrade defects. The experimental results indicates that compared with traditional SVM+HOG method and the baseline Faster R-CNN, the improved model can achieve better performance. The model robustness is demonstrated by a further comparison experiment of various defect types. In addition, the improvements to model performance of each improvement strategy are verified by an ablation experiment of improvement strategies. This paper tries to explore the new thinking for the application of deep learning method in the field of railway subgrade defect recognition.http://dx.doi.org/10.1155/2018/4832972 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Xinjun Xu Yang Lei Feng Yang |
spellingShingle |
Xinjun Xu Yang Lei Feng Yang Railway Subgrade Defect Automatic Recognition Method Based on Improved Faster R-CNN Scientific Programming |
author_facet |
Xinjun Xu Yang Lei Feng Yang |
author_sort |
Xinjun Xu |
title |
Railway Subgrade Defect Automatic Recognition Method Based on Improved Faster R-CNN |
title_short |
Railway Subgrade Defect Automatic Recognition Method Based on Improved Faster R-CNN |
title_full |
Railway Subgrade Defect Automatic Recognition Method Based on Improved Faster R-CNN |
title_fullStr |
Railway Subgrade Defect Automatic Recognition Method Based on Improved Faster R-CNN |
title_full_unstemmed |
Railway Subgrade Defect Automatic Recognition Method Based on Improved Faster R-CNN |
title_sort |
railway subgrade defect automatic recognition method based on improved faster r-cnn |
publisher |
Hindawi Limited |
series |
Scientific Programming |
issn |
1058-9244 1875-919X |
publishDate |
2018-01-01 |
description |
Railway subgrade defect is the serious threat to train safety. Vehicle-borne GPR method has become the main railway subgrade detection technology with its advantages of rapidness and nondestructiveness. However, due to the large amount of detection data and the variety in defect shape and size, defect recognition is a challenging task. In this work, the method based on deep learning is proposed to recognize defects from the ground penetrating radar (GPR) profile of subgrade detection data. Based on the Faster R-CNN framework, the improvement strategies of feature cascade, adversarial spatial dropout network (ASDN), Soft-NMS, and data augmentation have been integrated to improve recognition accuracy, according to the characteristics of subgrade defects. The experimental results indicates that compared with traditional SVM+HOG method and the baseline Faster R-CNN, the improved model can achieve better performance. The model robustness is demonstrated by a further comparison experiment of various defect types. In addition, the improvements to model performance of each improvement strategy are verified by an ablation experiment of improvement strategies. This paper tries to explore the new thinking for the application of deep learning method in the field of railway subgrade defect recognition. |
url |
http://dx.doi.org/10.1155/2018/4832972 |
work_keys_str_mv |
AT xinjunxu railwaysubgradedefectautomaticrecognitionmethodbasedonimprovedfasterrcnn AT yanglei railwaysubgradedefectautomaticrecognitionmethodbasedonimprovedfasterrcnn AT fengyang railwaysubgradedefectautomaticrecognitionmethodbasedonimprovedfasterrcnn |
_version_ |
1721335955791544320 |