Reinforcing Bar Segmentation from Depth-Camera-Captured Point Cloud Data

Reinforced concrete, a composite material composed of concrete and rebars, is one of the most widely used materials in buildings.Rebar inspection at construction sites is recommended to mitigate potential risks associated with the omission or improper installation of rebars.However, current rebar inspection conducted on-site is predominantly manual, which is time consuming and labor-intensive.In this study, a new method is proposed to segment the rebar-related point cloud data from depthcamera-captured point cloud data.This method utilizes the normal vectors of tangent planes at each point to segment the rebar-related portions leveraging the directional differences between rebar and floor-related points for effective segmentation.To validate this method, experiments were conducted with six 1.2 m rebars under various experimental conditions.The accuracy of segmentation was assessed by comparing the actual spacing between the rebars and the distances between the segmented rebar points.The results demonstrated that the proposed method can effectively segment rebars and accurately measure the interval instances.