Image-based bolt-loosening detection using an improved homography-based perspective rectification method

In this paper, a new perspective rectification method for image of bolts in flange joint is proposed and the image-based bolt-loosening detection process for the flange joint is established. The presented method can correct the perspective distortion for the picture using the four center points of c...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of civil structural health monitoring 2024-03, Vol.14 (3), p.513-526
Hauptverfasser: Luo, Jun, Zhao, Jie, Xie, ChengQian, Sun, Yi, Liu, Xinpeng, Yan, Zhitao
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In this paper, a new perspective rectification method for image of bolts in flange joint is proposed and the image-based bolt-loosening detection process for the flange joint is established. The presented method can correct the perspective distortion for the picture using the four center points of corner bolts in bolt-group connection. However, it would be difficult to find the four known points when the picture comes from a flange joint, because of the occlusion of component and the definition of picture. In this paper, a new homography-based perspective rectification method is proposed to correct the perspective distortion for image of each nut based on the identified corners of the gasket. The method to determine the corners of the partially obscured gasket is also a novelty of this paper. Based on the proposed perspective rectification method, the image-based bolt-loosening detection process for the flange joint is proposed. The proposed method is verified using a bolt connection in laboratory. The results show that the proposed method can effectively correct the perspective distortion for image of each nut and identify the bolt-loosening of the flange joint.
ISSN:2190-5452
2190-5479
DOI:10.1007/s13349-023-00722-4