Indoor Point Cloud Segmentation Using a Modified Region Growing Algorithm and Accurate Normal Estimation

With the development of 3D sensors, 3D point cloud data can now be obtained conveniently, which has made the development of automatic point cloud data processing technology crucial. Region growing is a commonly used algorithm to segment point cloud, which greatly depends on the accuracy of point normals and requires the tuning of two thresholds; namely, the increment threshold of curvature (σ _th ) and normal angles (θ _th ). In this paper, we improve the region growing algorithm in two ways: Accurate normal estimation and strengthening the region growing criteria. For the first aspect, principal component analysis (PCA) is utilized to estimate the initial normals of the point cloud. Then, the points are divided into regular points (RP) and sharp feature points (SFP), according to their initial normals. A robust estimator-based PCA is then applied to refine the SFP normals. For the latter aspect, non-connective points are detected according to shared neighbor points, and non-coplanar points are determined by comparing the residual with a robust scale of it. In addition, σ _th is set as the 95 ^th percentile of curvature, allowing for easier parameter adjustment. Finally, the segmentation effect of the proposed method is evaluated through internal and external indices. The results indicate that the proposed method can accurately estimate the point normals within an acceptable time, and can obtain a better segmentation result than the classic PCA-based region growing algorithm and advanced DetMM-based methods.