Pointwise classification of mobile laser scanning point clouds of urban scenes using raw data

Abstract

Mobile laser scanning (MLS), which can quickly collect a high-resolution and highprecision point cloud of the surroundings of a vehicle, is an appealing technology for threedimensional (3D) urban scene analysis. In this regard, the classification of MLS point clouds is a common and core task. We focus on pointwise classification, in which each individual point is categorized into a specific class by applying a binary classifier involving a set of local features derived from the neighborhoods of the point. To speed up the neighbor search and enhance feature distinctiveness for pointwise classification, we exploit the topological and semantic information in the raw data acquired by light detection and ranging (LiDAR) and recorded in scan order. First, a two-dimensional (2D) scan grid for data indexing is recovered, and the relative 3D coordinates with respect to the LiDAR position are calculated. Subsequently, a set of local features is extracted using an efficient neighbor search method with a low computational complexity independent of the number of points in a point cloud. These features are further merged to produce a variety of binary classifiers for specific classes via a GentleBoost supervised learning algorithm combining decision trees. The experimental results on the Paris-rue-Cassette database demonstrate that the proposed approach outperforms the state-of-the-art methods with a 10% improvement in the F 1 score, whereas it uses simpler geometric features derived from a spherical neighborhood with a radius of 0.5 m. © The Authors. Published by SPIE under a Creative Commons Attribution 4.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

High throughput analysis of leaf chlorophyll content in sorghum using RGB, hyperspectral, and fluorescence imaging and sensor fusion

“…With the advancement in sensor technologies, high-throughput plant phenotyping (HTPP) has become more widely available to the research community, in particular the high-throughput imaging technology. Many researchers believe that image acquisition is no longer the challenge; the true bottleneck is in the analysis of thousands of plant images that are acquired in a short time window [ 29 , 42 ]. Image-based features extracted from images, which contain the information on the morphological and biochemical traits of plants, enable effective use of genomic data to bridge the genotype-to-phenotype gap for crop improvement.…”

Hierarchical SVM for Semantic Segmentation of 3D Point Clouds for Infrastructure Scenes

“…Various terms are employed in scholarly discussions to characterize point cloud segmentation and classification, including semantic segmentation, instance segmentation, point labeling, and point-wise classification [32][33][34]. Semantic segmentation involves assigning semantic labels based on classes, while instance segmentation goes further to distinguish individual instances or objects within the same class, providing a more detailed understanding of the scene.…”

Low-volume precision spray for plant pest control using profile variable rate spraying and ultrasonic detection

“…The importance of proper selection of spray application methods is being increasingly recognized in precision agricultural and forestry management. With the growing awareness of environmental conservation and public health, pesticide application requires accurate, efficient, scientific, and reasonable operation according to the actual conditions and circumstances ( Fang et al., 2021 ; Li et al, 2021 ). Effective agricultural and forestry diseases and insect pests control means not only maximizing the effectiveness of the pesticide application, but also minimizing the unintended effects (health hazard and environmental pollution).…”