Individualization of Pinus radiata Canopy from 3D UAV Dense Point Clouds Using Color Vegetation Indices

Abstract

The location of trees and the individualization of their canopies are important parameters to estimate diameter, height, and biomass, among other variables. The very high spatial resolution of UAV imagery supports these processes. A dense 3D point cloud is generated from RGB UAV images, which is used to obtain a digital elevation model (DEM). From this DEM, a canopy height model (CHM) is derived for individual tree identification. Although the results are satisfactory, the quality of this detection is reduced if the working area has a high density of vegetation. The objective of this study was to evaluate the use of color vegetation indices (CVI) in canopy individualization processes of Pinus radiata. UAV flights were carried out, and a 3D dense point cloud and an orthomosaic were obtained. Then, a CVI was applied to 3D point cloud to differentiate between vegetation and nonvegetation classes to obtain a DEM and a CHM. Subsequently, an automatic crown identification procedure was applied to the CHM. The results were evaluated by contrasting them with results of manual individual tree identification on the UAV orthomosaic and those obtained by applying a progressive triangulated irregular network to the 3D point cloud. The results obtained indicate that the color information of 3D point clouds is an alternative to support individualizing trees under conditions of high-density vegetation.

Development of a river DTM generation algorithm based on SfM point clouds using vegetation and morphological filters

“…It also faces difficulties in capturing accurate DTMs in areas with dense vegetation [18]. Methods for generating DTMs from point clouds can be broadly classified into morphological filtering methods [21][22][23][24][25][26][27][28][29], vegetation index-based methods [30][31][32][33][34], composite methods [35][36][37][38], and machine learning-based methods [39][40][41] [35] effectively removed vegetation in steep and rugged terrains using morphological filters such as Progressive Morphological Filter (PMF), Simple Morphological Filter (SMRF), Cloth Simulation Filter (CSF), Adaptive Triangular Irregular Network (ATIN), and CAractérisation de NUages de POints (CANUPO). [42] achieved vegetation removal using CSF and ATIN.…”

Image Segmentation and Filtering of Anaerobic Lagoon Floating Cover in Digital Elevation Model and Orthomosaics Using Unsupervised k-Means Clustering for Scum Association Analysis

“…Booth et al [14] achieved 97% classification accuracy rate in landslide mapping by using spectrum-based methods and filtering unwanted non-native features with the assumption that they exhibit higher spatial frequency. Considerable progress has been achieved in the active field of vegetation suppression within geospatial models, which includes colour-based and slope-based filtering techniques [15][16][17][18], with commercial software integrating proprietary algorithms specifically designed for vegetation filtering in ground terrain elevation [19]. However, in specialised applications, the necessity arises for employing highly specific processing strategies, including multiple stages of filtering [15,17,18,20,21].…”