.png){kind=icon}
.png){kind=icon}
Return
An automatic method for measuring inter-storey structural displacement based on vision data
.png){kind=icon}
PRE
AI
DOI：10.1016/j.ymssp.2025.113123.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Natural disasters, such as earthquakes, tsunamis, and storm surges, frequently cause significant structural damage, thus requiring detailed post-disaster assessments. Vibration-based structural damage detection requires accurate structural vibration signals. Compared to sensor-based displacement measurement method, vision-based method is more cost-effective, utilizing visual algorithms to extract displacement data from video recordings. This study proposes a Surveillance Camera-based Inter-storey Displacement Measurement System (SCIDMS) which utilizes urban surveillance cameras to monitor inter-storey displacements during disasters. However, standard surveillance footage is often low-resolution and high-noise during high vibrations due to unstable voltage and rapid motion. To address these challenges, an Improved Kernelized Correlation Filter (IKCF) algorithm is developed herein. The IKCF incorporates an image pyramid for expanded tracking range, edge-enhancing image operators, and Discrete Fourier Transform (DFT) up-sampling for sub-pixel accuracy. The proposed method is validated through a shaking table test on a full-scale multi-layer frame concrete structure and applied to real earthquake videos. The algorithm’s accuracy and effectiveness are compared with other approaches. The proposed IKCF method achieved the most accurate displacement results across various simulated degraded videos in the shaking table test and effectively captured the frequency-domain characteristics of inter-story displacements. For real earthquake videos, despite the absence of ground-truth displacement measurements, the IKCF method demonstrated the ability to extract the most reasonable displacement signal, free from outlier values, compared to other algorithms. This approach presents a cost-effective solution for SCIDMS in densely populated urban areas following disaster events.
Keywords:
surveillance camera
inter-storey displacement
damage detection
vision-based method
improved kernelized correlation filter
Journal
IF:
8.9
Papers:
1.3W
Citations:
6.6W
Organization
View more organizations