건축물 사후관리를 위한 포인트 클라우드 기반 표면균열 시뮬레이션 및 측정 알고리즘 개발

초록

Surface cracks in buildings are critical indicators of structural safety, and even minor defects may require immediate repair if they exceed regulatory thresholds. Such cracks can signify underlying structural flaws or construction deficiencies, and may also cause psychological discomfort and practical inconvenience for occupants. In particular, comprehensive inspection of apartment buildings is challenging due to the large number of units and extensive surface area. Traditional visual inspections conducted by skilled personnel rely heavily on human experience and are inherently subjective. To address this issue, this study proposes a parametric method for measuring crack widths and determining whether they meet the criteria for critical defects. A quantitative standard for crack evaluation is presented, with a focus on identifying major cracks classified as critical. A parametric mock-up was developed for algorithm creation and validation. To identify cracks exceeding the critical threshold of 0.3 mm, specifications for 3D scanners and point cloud density were derived, and wall surfaces with simulated cracks were generated accordingly. Crack-related points were extracted, and the distances between them were measured using a parametric algorithm to determine the width of the crack. The algorithm was then validated by comparing the results against the critical defect threshold. Preprocessing steps in the point cloud acquisition process were excluded in this study, and non-orthogonal architectural forms were not considered. Future research will aim to automate initial crack detection and preprocessing using deep learning techniques, enhancing the practical application of this approach in post-construction building maintenance.

키워드