Comparative analysis of crack information extraction using structured light and SfM techniques

Abstract

Structural safety is a critical concern in civil engineering, making the monitoring of pathological manifestations essential, particularly cracks, which are among the most frequent defects in concrete structures. Accurate crack characterization is fundamental to ensuring structural integrity and preventing potential failures. Traditional assessment methods often rely on destructive testing or visual inspection, which may either damage the structure or depend heavily on operator expertise. In contrast, non-destructive techniques enable more reliable and repeatable data acquisition. This study presents a comparative analysis of crack characterization using structured light and Structure from Motion (SfM) techniques. Both approaches generate point clouds from which geometric features, such as crack width, can be extracted. The case study was conducted on a crack located at the Botanic University Restaurant in Curitiba, Paraná, Brazil. Data acquisition was performed using a portable infrared structured light scanner and a smartphone for the SfM approach. Calibrated targets, referenced by a total station, were used as control points. The results demonstrate that both techniques are suitable for geometric characterization, although they present distinct performance in terms of accuracy and consistency. Structured light provided more stable local measurements, while SfM showed better agreement with reference distances for global geometry.