Short-distance and long-distance bridge displacement measurement based on template matching and feature detection methods

With the development of computers and image equipment, vision-based methods begin to be widely used in deformation measurement. However, whether in the short-term or the long-term measurement, the determination of the region of interest (ROI) mostly relies on adjustments with calculation results or subjective experience, which is difficult to meet the requirements of standardization in engineering. Therefore, this paper explores the influence of various sizes on the ROI and proposes the ideal size. Meanwhile, a vision-based approach combining template matching and feature point methods is proposed. The template matching method is used to determine the location of the target and the parameters of the ROI while the feature detection (FD) method is focused on extracting the displacement. Laboratory models, high-speed rail bridges and cable-stayed bridge experiments are conducted to verify the accuracy and stability of the proposed method. Furthermore, it is shown that millimeter-level displacement measurement can be achieved with short-distance and long-distance monitoring devices.