Video‐Based Micro‐Vibration Measurement for Hydraulic Structures in Field Environments Using Phase‐Based Motion Extraction

Hydraulic structures subjected to prolonged operational conditions inevitably experience damage and potential structural resonance, which threaten their stability and safety. Structural health monitoring (SHM) is a well‐established approach to ensuring safety; however, conventional methods primarily rely on contact sensors, which are inadequate for comprehensive monitoring of large and complex structures. Video‐based noncontact monitoring techniques effectively address this limitation. This study proposes a field‐deployable, phase‐based video method to measure micro‐vibrations of large hydraulic structures. Rather than batch processing full frames, vibration time histories are extracted from user‐defined pixels or pixel groups, enabling efficient region of interest (ROI)—level analysis with synchronized high‐speed videos. We then assess the method's robustness under varied illumination conditions, shooting angles, and scales, and visualize micro‐motions using frequency‐selective motion magnification. On an aging aqueduct, video‐derived responses show strong agreement with those obtained via displacement sensors; the correlation coefficient with sensor data reaches up to 0.96, accurately capturing the spectral peak bands within the 0−100 Hz and demonstrating exceptional robustness in the 0−20 Hz. Based on the discussion of practical limits, we provide guidelines for low‐texture, large‐scene monitoring. Adequate and uniform illumination is essential, and local fine‐scale features can significantly improve accuracy. The correlation coefficient can be increased by 0.1–0.3. Perpendicularity between the shooting angle and the target vibration direction is critical for obtaining high‐precision data.