LiDAR Data Fusion to Improve Forest Attribute Estimates: A Review

Current Forestry Reports, Journal Year: 2024, Volume and Issue: 10(4), P. 281 - 297

Published: June 21, 2024

Abstract Purpose of the Review Many LiDAR remote sensing studies over past decade promised data fusion as a potential avenue to increase accuracy, spatial-temporal resolution, and information extraction in final products. Here, we performed structured literature review analyze relevant on these topics published last main motivations applications for fusion, methods used. We discuss findings with panel experts report important lessons, challenges, future directions. Recent Findings other datasets, including multispectral, hyperspectral, radar, is found be useful variety literature, both at individual tree level area level, tree/crown segmentation, aboveground biomass assessments, canopy height, species identification, structural parameters, fuel load assessments etc. In most cases, gains are achieved improving accuracy (e.g. better classifications), resolution height). However, questions remain regarding whether marginal improvements reported range worth extra investment, specifically from an operational point view. also provide clear definition “data fusion” inform scientific community combination, integration. Summary This provides positive outlook come, while raising about trade-off between benefits versus time effort needed collecting combining multiple datasets.

Language: Английский

---

Geomatic Data Fusion for 3D Tree Modeling: The Case Study of Monumental Chestnut Trees

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(8), P. 2197 - 2197

Published: April 21, 2023

In recent years, advancements in remote and proximal sensing technology have driven innovation environmental land surveys. The integration of various geomatics devices, such as reflex UAVs equipped with RGB cameras mobile laser scanners (MLS), allows detailed precise surveys monumental trees. With these data fusion method, we reconstructed three 3D tree models, allowing the computation metric variables diameter at breast height (DBH), total (TH), crown basal area (CBA), volume (CV) wood (WV), even providing information on shape its overall conditions. We processed point clouds software CloudCompare, Forest, R MATLAB, whereas photogrammetric processing was conducted Agisoft Metashape. Three-dimensional models enhance accessibility to allow for a wide range potential applications, including development model (TIM), monitoring health, growth, biomass carbon sequestration. encouraging results provide basis extending virtualization trees larger scale conservation monitoring.

Language: Английский

---

Using Pre-Fire High Point Cloud Density LiDAR Data to Predict Fire Severity in Central Portugal

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(3), P. 768 - 768

Published: Jan. 29, 2023

The wall-to-wall prediction of fuel structural characteristics conducive to high fire severity is essential provide integrated insights for implementing pre-fire management strategies designed mitigate the most harmful ecological effects in fire-prone plant communities. Here, we evaluate potential point cloud density LiDAR data from Portuguese áGiLTerFoRus project characterize surface and canopy structure predict wildfire severity. study area corresponds a pilot flight around 21,000 ha central Portugal intersected by mixed-severity that occurred one month after survey. Fire was assessed through differenced Normalized Burn Ratio (dNBR) index computed pre- post-fire Sentinel-2A Level 2A scenes. In addition continuous data, also categorized (low or high) using appropriate dNBR thresholds communities area. We several metrics related distribution fuels strata with mean 10.9 m−2. Random Forest (RF) algorithm used capacity set accuracy RF regression classification model respectively, remarkably (pseudo-R2 = 0.57 overall 81%) considering only focused on variables loading. highest contribution models were proxies horizontal continuity (fractional cover metric) loads openness up 10 m height (density metrics), indicating increased higher load vertical continuity. Results evidence technical specifications acquisitions framed within enable accurate predictions density.

Language: Английский

---

Revealing Three-Dimensional Variations in Fuel Structures in Subtropical Forests through Backpack Laser Scanning

Forests, Journal Year: 2024, Volume and Issue: 15(1), P. 155 - 155

Published: Jan. 11, 2024

Wildfire hazard is a prominent issue in subtropical forests as climate change and extreme drought events increase frequency. Stand-level fuel load forest structure are determinants of fire occurrence spread. However, current management often lacks detailed vertical distribution, limiting accurate risk assessment effective policy implementation. In this study, backpack laser scanning (BLS) used to estimate several 3D structural parameters, including canopy height, crown base volume, stand density, vegetation area index (VAI), coverage, characterize the characteristics density distribution variation different stands China. Through standard measurement using BLS point cloud data, we found that VAI lower middle-height strata differed significantly among types. Compared LiDAR-derived can better show significant stratified changes direction Among types, conifer-broadleaf mixed C. lanceolata had higher surface than other while P. massoniana were particularly unique having strata, corresponding ladder stand, respectively. To provide more informative support for management, LiDAR data combined with remote sensing advocated facilitate visualization development assessment.

Language: Английский

---

Classification and Mapping of Fuels in Mediterranean Forest Landscapes Using a UAV-LiDAR System and Integration Possibilities with Handheld Mobile Laser Scanner Systems

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(18), P. 3536 - 3536

Published: Sept. 23, 2024

In this study, we evaluated the capability of an unmanned aerial vehicle with a LiDAR sensor (UAV-LiDAR) to classify and map fuel types based on Prometheus classification in Mediterranean environments. UAV data were collected across 73 forest plots located NE Spain. Furthermore, from handheld mobile laser scanner system (HMLS) 43 out used assess extent improvement identification resulting fusion HMLS data. three-dimensional point clouds (average density: 452 points/m2) allowed generation metrics indices related vegetation structure. Additionally, voxels 5 cm3 derived 63,148 facilitated calculation volume at each type height stratum (0.60, 2, 4 m). Two different models three machine learning techniques (Random Forest, Linear Support Vector Machine, Radial Machine) employed types: one including only variables other incorporating The most relevant introduced into models, according Dunn’s test, 99th 10th percentile heights, standard deviation total returns above m, Height Diversity Index (LHDI). best using was achieved Random Forest (overall accuracy = 81.28%), confusion mainly found between similar shrub tree types. integration yielded substantial improvement, especially 95.05%). mapping model correctly estimated area 55 least part 59 plots. These results confirm that UAV-LiDAR systems are valid operational tools for show how refines types, contributing more effective management ecosystems.

Language: Английский

---

Modelling internal tree attributes for breeding applications in Douglas-fir progeny trials using RPAS-ALS

Science of Remote Sensing, Journal Year: 2022, Volume and Issue: 7, P. 100072 - 100072

Published: Dec. 27, 2022

Coastal Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) is one of the most commercially important softwood species in North America. In British Columbia, Canada, breeding has increased volume gains between 20 and 30%, while 97% seedlings come from improved seed sources. Branching traits particular, have a strong influence on strength stiffness wood; however, they are rarely measured. Remotely Piloted Aerial Systems Airborne Laser Scanning (RPAS-LS) produce high-density three-dimensional point clouds that can be used for creation internal geometric features describing individual tree branching structures. We analyzed progeny test trial located developed new method to estimate branch attributes RPAS-LS data inclusion as selection criteria improvement programs. Branch length, angle, width, were estimated each tree. Narrow-sense heritability (the proportion variation due genetics) genetic correlations also estimated. The extracted length with correlation (r) 0.93 compared manual measurements. Using these attributes, results then show angle had highest (0.277), height (0.668). These findings encouraging forest managers indicate level metrics should considered when selecting trees

Language: Английский

---

Tree Branch Characterisation from Point Clouds: a Comprehensive Review

Current Forestry Reports, Journal Year: 2024, Volume and Issue: 10(5), P. 360 - 385

Published: July 25, 2024

Abstract Purpose of Review Since the late 1990s, researchers have been increasingly utilising digital methodologies to assess branch structure trees. The emergence commercial terrestrial laser scanners during this period catalysed an entirely new domain focused on point cloud-based research. Over years, field has transformed from a complex computational discipline into practical tool that effectively supports research endeavours. Through combined use non-destructive remote sensing techniques and advanced analytical methods, characterisation can now be carried out at unprecedented level. Recent Findings While scanning traditionally dominant methodology for domain, increased mobile unmanned aerial vehicles indicates transition towards more platforms. Quantitative structural modelling (QSM) pivotal in advancing field, enhancing capabilities across diverse fields. past five years seen uptake 2D 3D deep learning as alternatives. Summary This article presents comprehensive synthesis approximately 25 characterisation, reviewing data capture technologies along with forest types tree species which these applied. It explores current trends dynamic research, gaps some key challenges remain within field. In review, we placed particular emphasis potential resolution significant challenge associated occlusion through utilisation technologies, such vehicles. We highlight need cohesive method assessing cloud quality derived model accuracy, benchmarking sets used test existing algorithms.

Language: Английский

---

Remote Sensing and GIS Applications in Wildfires

IntechOpen eBooks, Journal Year: 2023, Volume and Issue: unknown

Published: Oct. 24, 2023

Wildfires are closely associated with human activities and global climate change, but they also affect health, safety, the eco-environment. The ability of understanding wildfire dynamics is important for managing effects wildfires on infrastructures natural environments. Geospatial technologies (remote sensing GIS) provide a means to study at multiple temporal spatial scales using an efficient quantitative method. This chapter presents overview applications geospatial in management. Applications related pre-fire conditions management (fire hazard mapping, fire risk fuel mapping), monitoring detection, detection hot-spots, thermal parameters, etc.) post-fire condition (burnt area burn severity, soil erosion assessments, vegetation recovery assessments monitoring) discussed. Emphasis given roles multispectral sensors, lidar evolving UAV/drone processing, combining various environmental characteristics wildfires. Current previous researches presented, future research trends It wildly accepted that low-cost, multi-temporal conducting local, regional global-scale research, assessments.

Language: Английский

---

Fusion of airborne multimodal point clouds for vegetation parameter correction extraction in burned areas

Optics Express, Journal Year: 2024, Volume and Issue: 32(6), P. 8580 - 8580

Published: Feb. 5, 2024

Most experimental studies use unimodal data for processing, the RGB image point cloud cannot separate shrub and tree layers according to visible vegetation index, airborne laser is difficult distinguish between ground grass ranges, address above problems, a multi-band information fusing LiDAR constructed. In this study, collected from UAV platforms, including clouds clouds, were used construct fine canopy height model (using data) high-definition digital orthophotos data), fused with (CHM) by selecting Difference Enhancement Vegetation Index (DEVI) Normalised Green-Blue Discrepancy (NGBDI) after comparing accuracy of different indices. Morphological reconstruction CHM + DEVI/NGBDI fusion image, remove unreasonable values; training samples, using classification regression algorithm, segmentation range burned areas adaptive extraction as trees, shrubs grasslands, foreground markers local maximum algorithm detect apexes, non-tree are assigned be background markers, Watershed Transform performed obtain contour; original divided into chunks segmented single-tree contour, highest traversed search point, corrected elevations one one. Accuracy analysis extracted method measured showed that improved increased overall recall rate 4.1%, precision 3.7%, F1 score 3.9%, 8.8%, 1.4%, 1.7%, 6.4%, 1.8%, 0.3%, respectively, in six sampling plots. The effectiveness verified, while higher degree mixing region better effect algorithm.

Language: Английский

---

Multi-channel depth segmentation network based on 3D graph convolution algorithm and its application in point cloud segmentation

Alexandria Engineering Journal, Journal Year: 2024, Volume and Issue: 109, P. 740 - 753

Published: Sept. 30, 2024

Language: Английский

---