PreprintArticleVersion 1Preserved in Portico This version is not peer-reviewed
Automatic Semantic Segmentation of Salient Patterns in Temporal Images for Digital Image Correlation-based Structural Health Monitoring of Large Structures
Version 1
: Received: 20 May 2022 / Approved: 23 May 2022 / Online: 23 May 2022 (10:38:16 CEST)
How to cite:
Chan, C.C.K.; Kumar, D.; Chiang, C.; Ferworn, A. Automatic Semantic Segmentation of Salient Patterns in Temporal Images for Digital Image Correlation-based Structural Health Monitoring of Large Structures. Preprints2022, 2022050298. https://doi.org/10.20944/preprints202205.0298.v1
Chan, C.C.K.; Kumar, D.; Chiang, C.; Ferworn, A. Automatic Semantic Segmentation of Salient Patterns in Temporal Images for Digital Image Correlation-based Structural Health Monitoring of Large Structures. Preprints 2022, 2022050298. https://doi.org/10.20944/preprints202205.0298.v1
Chan, C.C.K.; Kumar, D.; Chiang, C.; Ferworn, A. Automatic Semantic Segmentation of Salient Patterns in Temporal Images for Digital Image Correlation-based Structural Health Monitoring of Large Structures. Preprints2022, 2022050298. https://doi.org/10.20944/preprints202205.0298.v1
APA Style
Chan, C.C.K., Kumar, D., Chiang, C., & Ferworn, A. (2022). Automatic Semantic Segmentation of Salient Patterns in Temporal Images for Digital Image Correlation-based Structural Health Monitoring of Large Structures. Preprints. https://doi.org/10.20944/preprints202205.0298.v1
Chicago/Turabian Style
Chan, C.C.K., Chih-Hung Chiang and Alexander Ferworn. 2022 "Automatic Semantic Segmentation of Salient Patterns in Temporal Images for Digital Image Correlation-based Structural Health Monitoring of Large Structures" Preprints. https://doi.org/10.20944/preprints202205.0298.v1
Abstract
Large structures such as wind turbines are subject to environmental factors and varying operational loads which may result in structural damage, making components of these large structures prone to performance and mechanical degradation. The use of high-definition optical vision sensors in digital image correlation (DIC) allow for the application of a non-destructive image registration technique in which it measures finite three-dimensional deformations on surfaces through correlations of a unique pattern or set of unique localized patterns. However, the physical placement of an artificial marker such as a unique speckled pattern on the surface of the structure is time-consuming and often impractical for large structures. Therefore, we propose a novel auto-mated methodology that searches and segments salient and unique regions of an image as well as for all subsequent images to assist in performing efficient displacement measurements for vibrational study and structural health monitoring purposes. Our algorithm is validated on a con-trolled set of images, as well as on a small structure and large real-world wind turbine, which suggests the algorithm’s efficacy without the use of artificial markers for large structural health monitoring.
Keywords
digital image correlation; semantic filter; structural health monitoring; unique salient patterns; wind turbine
Subject
Engineering, Civil Engineering
Copyright:
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.