A Supervoxel-Based Random Forest Method for Robust and Effective Airborne LiDAR Point Cloud Classification

Abstract

As an essential part of point cloud processing, autonomous classification is conventionally used in various multifaceted scenes and non-regular point distributions. State-of-the-art point cloud classification methods mostly process raw point clouds, using a single point as the basic unit and calculating point cloud features by searching local neighbors via the k-neighborhood method. Such methods tend to be computationally inefficient and have difficulty obtaining accurate feature descriptions due to inappropriate neighborhood selection. In this paper, we propose a robust and effective point cloud classification approach that integrates point cloud supervoxels and their locally convex connected patches into a random forest classifier, which effectively improves the point cloud feature calculation accuracy and reduces the computational cost. Considering the different types of point cloud feature descriptions, we divide features into three categories (point-based, eigen-based, and grid-based) and accordingly design three distinct feature calculation strategies to improve feature reliability. Two International Society of Photogrammetry and Remote Sensing benchmark tests show that the proposed method achieves state-of-the-art performance, with average F1-scores of 89.16 and 83.58, respectively. The successful classification of point clouds with great variation in elevation also demonstrates the reliability of the proposed method in challenging scenes.

Machine Learning and 3D Reconstruction of Materials Surface for Nondestructive Inspection

“…The extracted point cloud data were used to detect and classify weld defects at the level of their morphological characteristics. For this, a modified method was used with further analysis based on the random forest model [39][40][41]. We implemented and developed the procedure for the three-dimensional reconstruction of the surface of the pipeline weld.…”

Weakly supervised MS-LiDAR point cloud classification based on kernel convolutional semantic query network

“…In the classification regression part of the network, the spatial residual attention mechanism algorithm is used to form a specific feature for the MS-LiDAR point cloud, simplify the processing of the attention mechanism [9][10], integrate all the feature maps, and carry out average pooling processing at the same time, so as to complete the inference calculation of the multi-label topological relationship T , as shown in formula (2):…”

A Method for Tomato Plant Stem and Leaf Segmentation and Phenotypic Extraction Based on Skeleton Extraction and Supervoxel Clustering

“…However, when the leaves overlap and adhere to each other, there is a situation where two different leaves are grouped together (Figure 6b). By observing that there are more obvious concave and convex mutations between overlapping and adhering leaves, a segmentation method based on supervoxel clustering [38] is used for further segmentation. The point cloud to be segmented is first subjected to supervoxel processing to obtain a supervoxel neighborhood map (Figure 6c).…”