Automating Parameter Learning for Classifying Terrestrial LiDAR Point Cloud Using 2D Land Cover Maps

Abstract

Abstract:The automating classification of point clouds capturing urban scenes is critical for supporting applications that demand three-dimensional (3D) models. Achieving this goal, however, is met with challenges because of the varying densities of the point clouds and the complexity of the 3D data. In order to increase the level of automation in the point cloud classification, this study proposes a segment-based parameter learning method that incorporates a two-dimensional (2D) land cover map, in which a strategy of fusing the 2D land cover map and the 3D points is first adopted to create labelled samples, and a formalized procedure is then implemented to automatically learn the following parameters of point cloud classification: the optimal scale of the neighborhood for segmentation, optimal feature set, and the training classifier. It comprises four main steps, namely: (1) point cloud segmentation; (2) sample selection; (3) optimal feature set selection; and (4) point cloud classification. Three datasets containing the point cloud data were used in this study to validate the efficiency of the proposed method. The first two datasets cover two areas of the National University of Singapore (NUS) campus while the third dataset is a widely used benchmark point cloud dataset of Oakland, Pennsylvania. The classification parameters were learned from the first dataset consisting of a terrestrial laser-scanning data and a 2D land cover map, and were subsequently used to classify both of the NUS datasets. The evaluation of the classification results showed overall accuracies of 94.07% and 91.13%, respectively, indicating that the transition of the knowledge learned from one dataset to another was satisfactory. The classification of the Oakland dataset achieved an overall accuracy of 97.08%, which further verified the transferability of the proposed approach. An experiment of the point-based classification was also conducted on the first dataset and the result was compared to that of the segment-based classification. The evaluation revealed that the overall accuracy of the segment-based classification is indeed higher than that of the point-based classification, demonstrating the advantage of the segment-based approaches.

A Hierarchical Approach for Point Cloud Classification With 3D Contextual Features

“…For the definition of neighborhood for points, this study adopts k-nearest-neighborhood (KNN), which is based on a fixed number of nearest points and approximates the density-adaptive search in an unevenly distributed point cloud, because point clouds are characterized by varying point densities. To identify the optimal number of nearest points k, the two-step approach proposed by [15] is employed. In the first step, a series of segmentation with different k values are conducted.…”

“…We also find that much of the obtained information from the 3D contextual features concurs with commonsense knowledge. Although existing studies have attempted to adopt such heuristics for classifying point clouds (e.g., [15]), the presented approach extracts the information automatically and utilizes it quantitatively. The 3D contextual information mined by the proposed method is also more comprehensive and robust.…”

“…The former, despite its importance, received minor attention because most published studies verify their performances using benchmark point cloud where labels of training data are often available. An exception is the study in Feng and Guo [15] where an automatic method for sample selection with the aid of 2D land cover maps and a built topological graph is provided. For the latter, spectral, geometrical, and eigen-based features are commonly used [7], [15].…”

“…An exception is the study in Feng and Guo [15] where an automatic method for sample selection with the aid of 2D land cover maps and a built topological graph is provided. For the latter, spectral, geometrical, and eigen-based features are commonly used [7], [15]. Both point-and segment-based approaches extract eigenbased features from a 3D covariance matrix, which comprises 3D coordinates of a cluster of points, to represent local characteristics of points.…”

Self-Learning Ontology for Instance Segmentation of 3d Indoor Point Cloud

“…We sort eigenvalues 1 , 2 , 3 , such as 1 > 2 > 3 , where linked eigen vectors 1 ⃗⃗⃗⃗ , 2 ⃗⃗⃗⃗ , 3 ⃗⃗⃗⃗ , respectively, represent the principal direction, its orthogonal direction, and the estimated plane normal. These indicators, as reviewed in Section 2, are interesting for deriving several eigen-based features (Feng and Guo, 2018), from which we use the omnivariance, planarity and verticality for their good informative description as seen in (Florent Poux et al, 2017;Poux and Billen, 2019a).…”