A Crown Morphology-Based Approach to Individual Tree Detection in Subtropical Mixed Broadleaf Urban Forests Using UAV LiDAR Data

Xu, Wenbing; Deng, Shengyuan; Liang, Dan

doi:10.3390/rs13071278

Cited by 19 publications

(14 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is because image analysis techniques generally produce better results on conifers due to their excurrent form and apical dominance, rather than on broadleaf species with their decurrent form with multiple leaders [17,18]. Some studies have tried to improve the performance of methods like local maximum filtering, watershed, and region-growing by applying multiscale analysis [19,20] or by integrating morphological features [21,22], but generalising those models to different scenes is timeconsuming as it requires further analysis on forest features and manual adjustments on parameters.…”

Section: Introductionmentioning

confidence: 99%

A Systematic Review of Individual Tree Crown Detection and Delineation with Convolutional Neural Networks (CNN)

et al. 2023

View full text Add to dashboard Cite

Purpose of Review Crown detection and measurement at the individual tree level provide detailed information for accurate forest management. To efficiently acquire such information, approaches to conduct individual tree detection and crown delineation (ITDCD) using remotely sensed data have been proposed. In recent years, deep learning, specifically convolutional neural networks (CNN), has shown potential in this field. This article provides a systematic review of the studies that used CNN for ITDCD and identifies major trends and research gaps across six perspectives: accuracy assessment methods, data types, platforms and resolutions, forest environments, CNN models, and training strategies and techniques. Recent Findings CNN models were mostly applied to high-resolution red–green–blue (RGB) images. When compared with other state-of-the-art approaches, CNN models showed significant improvements in accuracy. One study reported an increase in detection accuracy of over 11%, while two studies reported increases in F1-score of over 16%. However, model performance varied across different forest environments and data types. Several factors including data scarcity, model selection, and training approaches affected ITDCD results. Summary Future studies could (1) explore data fusion approaches to take advantage of the characteristics of different types of remote sensing data, (2) further improve data efficiency with customised sample approaches and synthetic samples, (3) explore the potential of smaller CNN models and compare their learning efficiency with commonly used models, and (4) evaluate impacts of pre-training and parameter tunings.

show abstract

Section: Introductionmentioning

confidence: 99%

A Systematic Review of Individual Tree Crown Detection and Delineation with Convolutional Neural Networks (CNN)

et al. 2023

View full text Add to dashboard Cite

show abstract

“…This method is based on the assumption that a treetop is the highest point within a crown and that the crown boundary is relatively low (Zhen et al, 2016;Xu et al, 2021). A pixel in CHM is identified as an LM when its neighbouring pixels have a lower height value (Koch et al, 2006;Plakman et al, 2020).…”

Section: Ethical Statementmentioning

confidence: 99%

“…The predicted living tree densities were based on the treetops of dominant trees, which were detected using fixed-window-size local-maxima (LM) filtering (Popescu and Wynne, 2004; Gebreslasie et al, 2011) applied to the Canopy Height Model (CHM) acquired from Light Detection and Ranging (LiDAR) data (obtained from airborne laser scanning with a density of 4 points per m 2 ; https://www.geoportal.gov.pl/en/data/lidar-measurements-lidar). This method is based on the assumption that a treetop is the highest point within a crown and that the crown boundary is relatively low (Zhen et al, 2016; Xu et al, 2021). A pixel in CHM is identified as an LM when its neighbouring pixels have a lower height value (Koch et al, 2006; Plakman et al, 2020).…”

Section: Remote Data – Prediction Of Density Of Living Treesmentioning

confidence: 99%

Spatial distribution of Tree-related Microhabitats in a primeval mountain forest: from natural patterns to landscape planning and forest management recommendations

Przepióra,

Lewandowski,

Ciach

2024

Preprint

View full text Add to dashboard Cite

Tree-related Microhabitats (TreMs) are essential for sustaining forest biodiversity. Although TreMs represent ephemeral resources that are spread across the landscape, their spatial distribution within temperate forests remains poorly understood. To address this knowledge gap, we conducted a study on 90 sample plots (0.05 ha each) located in a primeval mountain European beech Fagus sylvatica-dominated forest (Bieszczady Mountains, Carpathians). We explored the TreM profile with its link to habitat characteristics and described the spatial distribution of TreM indices. We identified 61 TreM types, with a mean richness of 19.7 +-4.9 SD TreM types per plot, a mean density of 740.7 +-292.5 SD TreM-bearing trees ha-1 and a mean TreM diversity of 1.2 +-0.1 SD. The diameter and living status of trees (living vs dead standing tree) were correlated with TreM richness on an individual tree. The stand structure, i.e. density and/or basal area of living and/or dead standing trees, and topographic conditions, i.e. slope exposure, were correlated with the TreM richness, density and diversity recorded on a study plot. We found no relationship between TreM richness, density and diversity and the presence of canopy gaps, which indicates that the influence of small-scale disturbances on the TreM profile is limited. However, our analysis revealed a clustered spatial pattern of TreM indices, with TreM-rich habitat patches (hot-spots) covering ~20% of the forest. A moderate TreM richness, density and diversity dominated ~60% of the forest, while TreM-poor habitat patches (cold-spots) covered ~20%. Based on our findings, we advise the transfer of knowledge on the spatial distribution of TreMs from primeval to managed forests and advocate the '2:6:2' triad rule: to allocate 20% of forests as strictly protected areas, to dedicate 60% to low-intensity forest management with the retention of large living trees and all dead standing trees, and to use the remaining 20% for intensive timber production. To ensure the continuance of the majority of TreM types, ≥ 55 living trees ha-1 > 60 cm in diameter should be retained. Such an approach will maintain a rich and diverse TreM assemblage across a broad spatial scale, which in turn will support biodiversity conservation and ecosystem restoration in secondary or managed forests.

show abstract

“…Thus, the main sensors used in UAVs are the LIDAR and RGB cameras, which are widely employed in UGS investigation. Although LIDAR sensors perform well in urban vegetation inventorying [31][32][33], the disadvantages of high cost and complex operation limit their application at large scales. Compared with the LIDAR sensor, RGB sensors mounted on a UAV are more suitable for large scale vegetation parameter measurements.…”

Section: Introductionmentioning

confidence: 99%

Estimation of the Three-Dimension Green Volume Based on UAV RGB Images: A Case Study in YueYaTan Park in Kunming, China

Hong

Liu

et al. 2023

Forests

View full text Add to dashboard Cite

Three-dimension green volume (3DGV) is a quantitative index that measures the crown space occupied by growing plants. It is often used to evaluate the environmental and climatic benefits of urban green space (UGS). We proposed the Mean of neighboring pixels (MNP) algorithm based on unmanned aerial vehicle (UAV) RGB images to estimate the 3DGV in YueYaTan Park in Kunming, China. First, we mapped the vegetated area by the RF algorithm based on visible vegetation indices and texture features, which obtained a producer accuracy (PA) of 98.24% and a user accuracy (UA) of 97.68%. Second, the Canopy Height Mode (CHM) of the vegetated area was built by using the Digital Surface Model (DSM) and Digital Terrain Model (DTM), and the vegetation coverage in specific cells (1.6 m × 1.6 m) was calculated based on the vegetation map. Then, we used the Mean of neighboring pixels (MNP) algorithm to estimate 3DGV based on the cell area, canopy height, and vegetation coverage. Third, the 3DGV based on the MNP algorithm (3DGV_MNP), the Convex hull algorithm (3DGV_Con), and the Voxel algorithm (3DGV_Voxel) were compared with the 3DGV based on the field data (3DGV_FD). Our results indicate that the deviation of 3DGV_MNP for plots (Relative Bias = 15.18%, Relative RMSE = 19.63%) is less than 3DGV_Con (Relative Bias = 24.12%, Relative RMSE = 29.56%) and 3DGV_Voxel (Relative Bias = 30.77%, Relative RMSE = 37.49%). In addition, the deviation of 3DGV_MNP (Relative Bias = 17.31%, Relative RMSE = 19.94%) is also less than 3DGV_Con (Relative Bias = 24.19%, Relative RMSE = 25.77%), and 3DGV_Voxel (Relative Bias = 27.81%, Relative RMSE = 29.57%) for individual trees. Therefore, it is concluded that the 3DGV estimation can be realized by using the Neighboring pixels algorithm. Further, this method performed better than estimation based on tree detection in UGS. There was 377,223.21 m3 of 3DGV in YueYaTan Park. This study provides a rapid and effective method for 3DGV estimation based on UAV RGB images.

show abstract

A Crown Morphology-Based Approach to Individual Tree Detection in Subtropical Mixed Broadleaf Urban Forests Using UAV LiDAR Data

Cited by 19 publications

References 48 publications

A Systematic Review of Individual Tree Crown Detection and Delineation with Convolutional Neural Networks (CNN)

A Systematic Review of Individual Tree Crown Detection and Delineation with Convolutional Neural Networks (CNN)

Spatial distribution of Tree-related Microhabitats in a primeval mountain forest: from natural patterns to landscape planning and forest management recommendations

Estimation of the Three-Dimension Green Volume Based on UAV RGB Images: A Case Study in YueYaTan Park in Kunming, China

Contact Info

Product

Resources

About