Improved defect detection and geometric characterization of drainage pipes in YOLOv7 algorithm

doi:10.16731/j.cnki.1671-3133.2024.03.015

Abstract

Abstract: Regular inspection of drainage pipes can find serious defects in time,which is of great significance to ensure the healthy operation of the drainage system and the safety of the urban environment.Aiming at the difficulty of detecting low illumination and low resolution of lower drainage pipes,an improved drainage pipeline defect detection and geometric characterization method of YOLOv7 algorithm is proposed.Firstly,the Contrast-Limited Adaptive Histogram Equalization (CLAHE) image enhancement technique is used to improve the contrast and detail of the image,so as to improve the detection network′s ability to capture drainage pipe defects.Secondly,based on the design of Drop-CA and MC modules,the YOLOv7 algorithm is improved,so that the network can obtain the semantic information of shallow defects,reduce the false detection rate,and improve the classification and localization capabilities of the model.Finally,for the two serious defects of crack and fracture,a method is designed to quantitatively describe the geometric characteristics of the defect to evaluate the size of the defect.Experimental results show that the final average accuracy of the improved network model reaches 93.3 %,and the detection speed reaches 42.9 f/s. This method effectively improves the accuracy of defect detection and classification of drainage pipelines,and can effectively characterize the geometric characteristics of defects.

Key words: image enhancement, defect detection, YOLOv7 algorithm, Drop-CA, geometric features

CLC Number:

TP271

ZENG Fei, LI Bin, ZHOU Jian, FAN Jiangfeng. Improved defect detection and geometric characterization of drainage pipes in YOLOv7 algorithm[J]. Modern Manufacturing Engineering, 2024, 522(3): 110-118.

References

[1] 陆绮荣,丁昕,梁雅雯.基于改进YOLOX的地下排水管道缺陷识别算法[J].电子测量技术,2022,45(21):161-168.
[2] 高琳,曹建国.管道泄漏检测方法研究综述[J].现代制造工程,2022(2):154-162.
[3] YANG M D,SU T C.Automated diagnosis of sewer pipe defects based on machine learning approaches[J].Expert Systems with Applications,2008,35(3):1327-1337.
[4] HALFAWY M R,HENGMEECHAI J.Automated defect detection in sewer closed circuit television images using histograms of oriented gradients and support vector machine[J].Automation in Construction,2014,38:1-13.
[5] XIE Q,LI D,XU J,et al.Automatic detection and classification of sewer defects via hierarchical deep learning[J].IEEE Transactions on Automation Science and Engineering,2019,16(4):1836-1847.
[6] CHENG J C P,WANG M.Automated detection of sewer pipe defects in closed-circuit television images using deep learning techniques[J].Automation in Construction,2018,95:155-171.
[7] WANG C Y, BOCHKOVSKIY A, LIAO H Y M. YOLOv7:
Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[J].ArXiv Preprint ArXiv:2207.02696,2022.
[8] HOU Q,ZHOU D,FENG J.Coordinate attention for efficient mobile network design[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Virtual:IEEE&CVF,2021.
[9] CHEN J,KAO S,HE H,et al.Run,Don′t Walk:Chasing Higher FLOPS for Faster Neural Networks[J].ArXiv Preprint ArXiv:2303.03667,2023.
[10] TONG Z,CHEN Y,XU Z,et al.Wise-IoU:Bounding Box Regression Loss with Dynamic Focusing Mechanism[J].ArXiv Preprint ArXiv:2301.10051,2023.
[11] 毛清华,李世坤,胡鑫,等.基于改进YOLOv7的煤矿带式输送机异物识别[J].工矿自动化,2022,48(12):26-32.DOI:10.13272/j.issn.1671-251x.2022100011.
[12] LARSSON G,MAIRE M,SHAKHNAROVICH G.Fractalnet:Ultra-deep neural networks without residuals[J].ArXiv Preprint ArXiv:1605.07648,2016.
[13] XIAO J,ZHAO T,YAO Y,et al.Context augmentation and feature refinement network for tiny object detection[J].ICLR,2022:1-11.
[14] new-workspace-zyqyt.Storm drain model Dataset[EB/OL].Roboflow,2021[2023-06-17].https://universe.roboflow.com/new-workspace-zyqyt/storm-drain-model.
[15] WANG J L,DENG Y L,LI Y,et al.A review on detection and defect identification of drainage pipeline[J].Sci.Technol.Eng,2020,33:13520-13528.