Ensuring proper reinforcement placement and conducting accurate rebar inspection are essential in preventing structural failures of reinforced concrete structures. In current practice, rebar diameter and spacing are primarily inspected visually using ...
Ensuring proper reinforcement placement and conducting accurate rebar inspection are essential in preventing structural failures of reinforced concrete structures. In current practice, rebar diameter and spacing are primarily inspected visually using tape measurements. However, as project scale increases, this manual process becomes time-consuming, labor-intensive, and poses potential safety risks. To address these issues, recent studies have explored the use of laser scanners to automate rebar inspection.
Most prior studies, however, have been limited to controlled laboratory environments and have not fully captured the complexities of actual construction sites. In real projects, reinforcement is distributed across complex site environments, arranged in multiple directions, and affected by bending, splicing, and occlusion caused by obstacles or overlapping rebars. Such conditions hinder the acquisition of ideal scan data, and existing methods have not sufficiently addressed these practical constraints.
This study proposes an integrated framework that detects rebars from terrestrial laser-scanned 3D point clouds and automatically estimates their diameters and spacing. To reduce computational costs, the point cloud is downsampled, and reinforcement planes are segmented through a two-stage process based on orientation and distance. Rebar segments are then extracted from each reinforcement plane and classified into vertical and horizontal rebars. To minimize data loss, points removed during preprocessing are restored by comparing their coordinates with those of the original point cloud.
For diameter estimation, each rebar segment is sliced along its axis, and cross-sectional shapes are extracted from the slices. An optimal radial band that minimizes noise influence is identified, and the final diameter is determined by selecting the most frequently predicted nominal size. For spacing estimation, the center points of rebar segments and their associated points are projected onto an axis perpendicular to the reinforcement direction. The distances between adjacent projected centers are calculated to derive rebar spacing.
The proposed framework was validated using point cloud data collected from two real construction sites, confirming its applicability in practical field conditions. By considering the realistic challenges of construction environments—including complex reinforcement layouts, bending, splicing, and occlusion—this study provides a robust automated rebar inspection framework. The method enables site engineers to perform automated rebar diameter and spacing inspection reliably and consistently using laser scanning, contributing to improved safety, reduced inspection time, and lower labor costs.