01515107

Component

In this study, a noncontact vision-based sensing method is proposed to measure the full-field inplane surface displacement of a bridge structure. To improve computational efficiency, a complete video sequence is divided into many subsequences. The proposed method includes five independent modules for profile initiation by a dynamic programming algorithm, profile tracking by a contour tracking algorithm, visual mark detection by speed-up-robust features, visual mark tracking by a three-stage data association algorithm, and displacement field construction by a Delaunay triangulation algorithm. This vision-based method is applied in various experimental settings. Both small-scale laboratory and full-scale field tests were conducted. Camera related parameters such as distance to object and frame rate were also investigated through lab tests. The obtained measurements were then compared with those from the conventional method such as linear variable differential transformers. It was demonstrated that the measurements using the proposed method are in good agreement with those from the conventional method.

This paper was sponsored by TRB committee AFF40 Field Testing and Nondestructive Evaluation (NDE) of Transportation Structures.

01503729

14-2256

English

13p

2014

Transportation Research Board 93rd Annual Meeting

Digital/other

Figures; Photos; References; Tables

Algorithms; Cameras; Field studies; Highway bridges; Remote sensing; Structural health monitoring; Visual perception

Bridges and other structures; Highways; Maintenance and Preservation; I60: Maintenance

Transportation Research Board Annual Meeting 2014 Paper #14-2256

TRIS, TRB, ATRI

Jan 27 2014 2:48PM