Optimized Dual-template Matching Method for Mobile Device-based Vibration Measurement in Sea Trials

Vibration measurement during sea trials is essential for evaluating ship performance and ensuring operational safety; however, conventional accelerometer-based methods require physical installation and are often impractical or unsafe in confined and elevated spaces. Non-contact approaches have therefore gained attention for their flexibility and ease of deployment. Image-based vibration measurement allows direct displacement extraction from video sequences and has been widely investigated as a non-intrusive, wide-area technique in recent years. Nevertheless, most existing systems rely on high-cost, high-resolution equipment, limiting their applicability in field environments. This study introduces a mobile device–based approach using image-based vibration measurement techniques to enable convenient and cost-effective vibration acquisition. Under real sea-trial conditions, handheld recording on unstable decks induces low-frequency drift in the measured signal, causing the target to gradually move outside the predefined Region of Interest (ROI) and making vibration measurement impossible. In addition, ROI variability caused by small mobile screens and unstable deck motion introduces background interference or loss of key motion information, leading to amplitude underestimation and reduced reliability of vibration measurement. In this study, a mobile device–based vibration measurement framework using optimized image-processing techniques for sea-trial and mobile environments is proposed. Real-time ROI optimization and Dual Template Matching (DTM) are implemented to minimize user-dependent variability and compensate for low-frequency drift. The proposed framework was examined through both laboratory and sea-trial experiments, demonstrating its effectiveness in enabling stable vibration measurement and mitigating environmental disturbances using a low-cost mobile platform.