Classification of Handheld Laser Scanning Tree Point Cloud Based on Different KNN Algorithms and Random Forest Algorithm

Wenshu Lin,Yongsheng Xu,Jinzhuo Wu,Weiwei Fan,Haoran Liu

doi:10.3390/f12030292

Abstract

Handheld mobile laser scanning (HMLS) can quickly acquire point cloud data, and has the potential to conduct forest inventory at the plot scale. Considering the problems associated with HMLS data such as large discreteness and difficulty in classification, different classification models were compared in order to realize efficient separation of stem, branch and leaf points from HMLS data. First, the HMLS point cloud was normalized and ground points were removed, then the neighboring points were identified according to three KNN algorithms and eight geometric features were constructed. On this basis, the random forest classifier was used to calculate feature importance and perform dataset training. Finally, the classification accuracy of different KNN algorithms-based models was evaluated. Results showed that the training sample classification accuracy based on the adaptive radius KNN algorithm was the highest (0.9659) among the three KNN algorithms, but its feature calculation time was also longer; The validation accuracy of two test sets was 0.9596 and 0.9201, respectively, which is acceptable, and the misclassification mainly occurred in the branch junction of the canopy. Therefore, the optimal classification model can effectively achieve the classification of stem, branch and leaf points from HMLS point cloud under the premise of comprehensive training.

Highlights

The geometric features of Handheld mobile laser scanning (HMLS)-based tree point cloud were calculated based on three kinds of K-Nearest neighbors (KNN) search methods
It is shown that regardless of the near-neighbor point search algorithm, the most relevant feature was the number of nearby points N, which can be regarded as the density parameter of the point cloud
Curvature Cλ, perpendicularity V and the ratio of eigenvalues for point cloud plane Rλ play an important role in point cloud classification, indicating that the spatial distribution characteristics of HMLS point clouds of stems, branches and leaves are quite different, which can be used as a basis for classification

Summary

Introduction

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. As a new type of ground-based mobile laser scanning (MLS) technology, Handheld. Mobile Laser Scanning (HMLS) can acquire 3D point cloud quickly and efficiently. It uses high-precision simultaneous localization and mapping (SLAM) algorithm to realize automatic splicing of point clouds, and the three-dimensional point cloud data of the scanned object can be obtained without GNSS signals [1]. In the forest environment, HMLS and other MLS methods (e.g., smart-phone, backpack, all-terrain vehicle, unmanned aircraft vehicle (UAV)-based laser scanning methods) are currently at research level and they are not yet applied operationally for field reference data collection [2,3]

Objectives

Methods

Results

Discussion

Conclusion