Answers are grounded in this paper and cite their sources.
The early detection of cracks is essential for maintaining structural integrity and reliability in various structures. Mode-shape-based crack detection has been widely investigated because cracks can be identified from damage-induced singularities or local changes in structural mode shapes. Wavelet-based methods have been investigated for crack detection, but the analysis at low spatial frequencies can suffer from broadened wavelet responses, reducing spatial resolution and diffusing crack localization. Image-based vibration measurement enables high-resolution mode shape estimation through dense, non-contact sensing; however, image-derived mode shapes may contain spatially inconsistent high-frequency noise arising from uncertainties inherent in image-based measurements. This study proposes a mode-shape-based crack localization framework that combines improved phase-based optical flow (POF) vibration extraction with a weighted multiplicative combination of multi-mode responses using a 1D Gabor filter bank. In the POF stage, modified phase unwrapping estimates phase variations with respect to the initial reference frame while tracking wrapping events, and magnitude filtering selects reliable pixels along structural boundaries. For crack localization, a crack-sensitive 1D Gabor filter bank is configured in the low-spatial-frequency range with an adjusted envelope width to improve spatial localization. The Gabor magnitudes from multiple modes are multiplicatively combined to emphasize consistent crack-induced anomalies while suppressing inconsistent image-based measurement errors. The proposed method was examined using cantilever beams with cracks of different sizes and locations and compared with CWT-based methods in terms of spatial concentration and localization error.
No takes yet. Share an insight, caveat, or question.
Eom et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: