A Method of Locating and Measuring Train Wheel Tread Defects Based on YOLOv3-tiny

Aiming at the problem of insufficient detection frame size accuracy in the deep learning based wheelset tread defect detection algorithm, this paper proposes a tread defect detection algorithm based on YOLOv3-tiny and traditional image algorithm, which can realize fast positioning of defects with lo...

Full description

Saved in:
Bibliographic Details
Main Authors: JIN Kairong, WANG Junping, CHEN Shenglan
Format: Article
Language:zho
Published: Editorial Office of Control and Information Technology 2022-04-01
Series:Kongzhi Yu Xinxi Jishu
Subjects:
Online Access:http://ctet.csrzic.com/thesisDetails#10.13889/j.issn.2096-5427.2022.02.200
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849224966129582080
author JIN Kairong
WANG Junping
CHEN Shenglan
author_facet JIN Kairong
WANG Junping
CHEN Shenglan
author_sort JIN Kairong
collection DOAJ
description Aiming at the problem of insufficient detection frame size accuracy in the deep learning based wheelset tread defect detection algorithm, this paper proposes a tread defect detection algorithm based on YOLOv3-tiny and traditional image algorithm, which can realize fast positioning of defects with low CPU consumption and precise measurement of geometric parameters. First, image enhancement is performed on the small samples obtained in the industrial scene, and then the YOLOv3-tiny algorithm is used to perform migration learning on tread defects to achieve rough localization of defects. In order to solve the key problems that the detection frame is too large and too small, traditional image algorithms such as Fourier transform, band-stop filter, and threshold segmentation are used to construct a defect size measurement model, contours of roughly located defects are extracted and detection frame size is optimized, and finally location and size of defect are accurately calculated. Defect location experimental results show that the average accuracy of defect identification is 89.4% and the CPU consumption does not exceed 10% when IoU threshold is 0.5. Experimental results of defect measurement show that the algorithm can optimize 74 out of 90 inspection frames and obtain a more accurate defect size. The above experimental results show that the detection algorithm in this paper is effective in improving size accuracy of detection frame.
format Article
id doaj-art-94b4c61b77074aa3a19e7994e09d26aa
institution Kabale University
issn 2096-5427
language zho
publishDate 2022-04-01
publisher Editorial Office of Control and Information Technology
record_format Article
series Kongzhi Yu Xinxi Jishu
spelling doaj-art-94b4c61b77074aa3a19e7994e09d26aa2025-08-25T06:48:02ZzhoEditorial Office of Control and Information TechnologyKongzhi Yu Xinxi Jishu2096-54272022-04-01697524252872A Method of Locating and Measuring Train Wheel Tread Defects Based on YOLOv3-tinyJIN KairongWANG JunpingCHEN ShenglanAiming at the problem of insufficient detection frame size accuracy in the deep learning based wheelset tread defect detection algorithm, this paper proposes a tread defect detection algorithm based on YOLOv3-tiny and traditional image algorithm, which can realize fast positioning of defects with low CPU consumption and precise measurement of geometric parameters. First, image enhancement is performed on the small samples obtained in the industrial scene, and then the YOLOv3-tiny algorithm is used to perform migration learning on tread defects to achieve rough localization of defects. In order to solve the key problems that the detection frame is too large and too small, traditional image algorithms such as Fourier transform, band-stop filter, and threshold segmentation are used to construct a defect size measurement model, contours of roughly located defects are extracted and detection frame size is optimized, and finally location and size of defect are accurately calculated. Defect location experimental results show that the average accuracy of defect identification is 89.4% and the CPU consumption does not exceed 10% when IoU threshold is 0.5. Experimental results of defect measurement show that the algorithm can optimize 74 out of 90 inspection frames and obtain a more accurate defect size. The above experimental results show that the detection algorithm in this paper is effective in improving size accuracy of detection frame.http://ctet.csrzic.com/thesisDetails#10.13889/j.issn.2096-5427.2022.02.200tread defectobject detectionYOLOv3-tinyFourier transformdeep learning
spellingShingle JIN Kairong
WANG Junping
CHEN Shenglan
A Method of Locating and Measuring Train Wheel Tread Defects Based on YOLOv3-tiny
Kongzhi Yu Xinxi Jishu
tread defect
object detection
YOLOv3-tiny
Fourier transform
deep learning
title A Method of Locating and Measuring Train Wheel Tread Defects Based on YOLOv3-tiny
title_full A Method of Locating and Measuring Train Wheel Tread Defects Based on YOLOv3-tiny
title_fullStr A Method of Locating and Measuring Train Wheel Tread Defects Based on YOLOv3-tiny
title_full_unstemmed A Method of Locating and Measuring Train Wheel Tread Defects Based on YOLOv3-tiny
title_short A Method of Locating and Measuring Train Wheel Tread Defects Based on YOLOv3-tiny
title_sort method of locating and measuring train wheel tread defects based on yolov3 tiny
topic tread defect
object detection
YOLOv3-tiny
Fourier transform
deep learning
url http://ctet.csrzic.com/thesisDetails#10.13889/j.issn.2096-5427.2022.02.200
work_keys_str_mv AT jinkairong amethodoflocatingandmeasuringtrainwheeltreaddefectsbasedonyolov3tiny
AT wangjunping amethodoflocatingandmeasuringtrainwheeltreaddefectsbasedonyolov3tiny
AT chenshenglan amethodoflocatingandmeasuringtrainwheeltreaddefectsbasedonyolov3tiny
AT jinkairong methodoflocatingandmeasuringtrainwheeltreaddefectsbasedonyolov3tiny
AT wangjunping methodoflocatingandmeasuringtrainwheeltreaddefectsbasedonyolov3tiny
AT chenshenglan methodoflocatingandmeasuringtrainwheeltreaddefectsbasedonyolov3tiny