A vision-based surface displacement/strain measurement technique based on robust edge-enhanced transform and algorithms for high spatial resolution

Longxi Luo*, Maria Q. Feng, Jianping Wu, Luzheng Bi

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

9 Citations (Scopus)
Plum Print visual indicator of research metrics
  • Citations
    • Citation Indexes: 8
  • Captures
    • Readers: 7
see details

Abstract

Measurement of two-dimensional surface displacement/strain distributions can be crucial in monitoring important structures. Computer vision techniques have the potentials for measuring surface displacements/strains. Conventional digital-image-correlation (DIC)-based computer vision techniques that have been applied in controlled conditions with artificially painted speckle patterns, however, have difficulties in robust measurement of structures' surface displacements against optical noises. Additionally, surface displacements obtained by DIC based on block-resolution template matching have limited accuracy because of low spatial resolutions and low level of smoothness. Therefore, a new computer vision technique SurfaceVision is proposed for accurate and robust surface displacement/strain measurement to tackle simulated field environmental conditions by incorporating multiple novel algorithms. First, a gradient-based edge-enhanced transform (EET) originally developed for one-point displacement tracking is extended for enabling robust surface displacement measurements against optical noises by manipulating gradient information rather than image intensities. Then, the improvement in the smoothness of surface displacements is enabled by incorporating EET with the iterative displacement optimization and the customized smart branching algorithms. Moreover, a novel pixel-resolution measurement algorithm is proposed for increasing the spatial resolution of surface displacements. Finally, an original intuitive strain conversion algorithm is developed for converting surface displacements into surface strains based on the principle similar to strain-gauge transducers. The performance of SurfaceVision is first validated in the numerical simulation and further demonstrated in the experiment of the four-point bending test. And a new method developed for predicting crack formations before they appear on structure surfaces, based on analyses of surface displacements/strains, is demonstrated.

Original languageEnglish
Article numbere2797
JournalStructural Control and Health Monitoring
Volume28
Issue number9
DOIs
Publication statusPublished - Sept 2021

Keywords

  • computer vision
  • edge-enhanced transform
  • smoothness
  • spatial resolution
  • surface displacements/strains

Fingerprint

Dive into the research topics of 'A vision-based surface displacement/strain measurement technique based on robust edge-enhanced transform and algorithms for high spatial resolution'. Together they form a unique fingerprint.

Cite this

Luo, L., Feng, M. Q., Wu, J., & Bi, L. (2021). A vision-based surface displacement/strain measurement technique based on robust edge-enhanced transform and algorithms for high spatial resolution. Structural Control and Health Monitoring, 28(9), Article e2797. https://doi.org/10.1002/stc.2797