Detection of the structural failing point using correlation and local correlation analysis

Construction worksites present numerous health and safety challenges. To address the challenges continuously monitoring the structural health, inspecting suspicious points at regular intervals, and promptly repairing the unstable structures play an important role in civil engineering as much as designing and building structures. This research aims to explore technologies for detecting failures in civil structures during construction. To apply image processing techniques and to facilitate the automated detection of structural failure points, we conducted a small-scale structural experiment for capturing chronological images taken from a video camera. This article examines three traditional and innovative image processing techniques: 1) count of pixels with changed intensities, 2) correlation coefficients, and 3) local correlation coefficients as robust image-based processing techniques. It was found that the local correlation method worked most efficiently, saving the computational time. As a future efficient usage, the local correlation method will be advantageous to developing possible potential algorithms for the automated detection system of failing points of civil structures in the civil engineering field.