Computer Vision-based Bridge Modal Identification: Equipment and Algorithms

Bridge modal identification based on computer vision is gradually becoming an emerging technical method for bridge health monitoring and disease detection. This paper reviews recent research literature on computer vision-based bridge modal identification, and systematically summarizes aspects including image acquisition equipment and image processing algorithms. The advantages, disadvantages, and applicable scenarios of common image acquisition devices such as industrial cameras, consumer cameras, and drones are analyzed. Industrial cameras offer good stability but require high technical skills from operators; drones are flexible and convenient but are easily affected by environmental factors like wind and rain; consumer cameras have low cost but limited monitoring accuracy. Image processing is a key step in computer vision-based bridge modal identification, and the selection of image processing algorithms plays a decisive role in the accuracy of result analysis. This paper focuses on the application of key algorithms such as digital image correlation (DIC) in bridge displacement extraction and modal identification. This study provides a systematic technical reference for dynamic displacement monitoring of bridge structures and discusses future development directions, which is of great significance for promoting intelligent and efficient bridge monitoring.