Full-waveform Airborne Laser Scanning Research Articles

Validation of beyond visual-line-of-sight drone photogrammetry for terrain and canopy height applications

Aerial drones enable detailed three-dimensional terrain and vegetation canopy reconstructions supporting a diverse set of environmental monitoring applications. Small drone-derived photogrammetric datasets (<5 km2) often extend field-scale observations and enable calibration and validation for satellite-based mapping. Applications usually require data over much larger areas or in remote areas such as the Arctic, yet large-area drone datasets are rare due to aviation regulations usually stipulating that missions adhere to within-visual-line-of-sight (VLOS) conditions. Terrain and vegetation canopy reconstruction based on longer-range drone missions requires evaluation to justify future beyond-visual-line-of-sight (BVLOS) data initiatives. This work analyzed BVLOS drone photogrammetric datasets for terrain and vegetation height applications along a forest-tundra ecotone (FTE) spanning 396 km in northern Canada. Specifically, we: 1) performed validation assessments of derived terrain models and canopy height models, and 2) investigated remote sensing applications supported by BVLOS data in a FTE experiencing the effects of climate change. The BVLOS data covered in excess of 550 km2, increasing conventional drone data coverage by two orders of magnitude. Elevation validation indicated that strong agreement with GNSS benchmarks, Airborne Laser Scanning (ALS) and VLOS drone data (0.15–0.19 m RMSE) is possible, and that vertical RMSEs <2x the image spatial resolution is achievable. To validate vegetation canopy height modelling capabilities, comparisons with NASA's Land, Vegetation, and Ice Sensor (LVIS) full-waveform ALS observations indicated satisfactory agreement for height metrics (R2 = 0.7 to 0.8, RMSEs <1.0 m). Canopies were reconstructed at fine spatial scales, which in turn supported a unique comparison of vegetation height distributions between six land-cover classes and ecoregions along the FTE, and identified discrepancies in how national and international canopy height models represented the FTE. Detected terrain subsidence, including retrogressive thaw slumping and ice-wedge degradation, illustrated how climate-driven permafrost thaw is modifying landscapes and placing infrastructure at increased risk, which requires intensive monitoring. These applications demonstrated how BVLOS-derived data enables broader scale assessments of regional vegetation structure and terrain changes across larger spatial extents than previously possible with VLOS drones. Mapping opportunities will arise in the environmental remote sensing discipline as drone systems are increasingly operated beyond visual range.

Towards Forest Condition Assessment: Evaluating Small-Footprint Full-Waveform Airborne Laser Scanning Data for Deriving Forest Structural and Compositional Metrics

Spatial data on forest structure, composition, regeneration and deadwood are required for informed assessment of forest condition and subsequent management decisions. Here, we estimate 27 forest metrics from small-footprint full-waveform airborne laser scanning (ALS) data using a random forest (RF) and automated variable selection (Boruta) approach. Modelling was conducted using leaf-off (April) and leaf-on (July) ALS data, both separately and combined. Field data from semi-natural deciduous and managed conifer plantation forests were used to generate the RF models. Based on NRMSE and NBias, overall model accuracies were good, with only two of the best 27 models having an NRMSE &gt; 30% and/or NBias &gt; 15% (Standing deadwood decay class and Number of sapling species). With the exception of the Simpson index of diversity for native trees, both NRMSE and NBias varied by less than ±4.5% points between leaf-on only, leaf-off only and combined leaf-on/leaf-off models per forest metric. However, whilst model performance was similar between ALS datasets, model composition was often very dissimilar in terms of input variables. RF models using leaf-on data showed a dominance of height variables, whilst leaf-off models had a dominance of width variables, reiterating that leaf-on and leaf-off ALS datasets capture different aspects of the forest and that structure and composition across the full vertical profile are highly inter-connected and therefore can be predicted equally well in different ways. A subset of 17 forest metrics was subsequently used to assess favourable conservation status (FCS), as a measure of forest condition. The most accurate RF models relevant to the 17 FCS indicator metrics were used to predict each forest metric across the field site and thresholds defining favourable conditions were applied. Binomial logistic regression was implemented to evaluate predicative accuracy probability relative to the thresholds, which varied from 0.73–0.98 area under the curve (AUC), where 11 of 17 metrics were &gt;0.8. This enabled an index of forest condition (FCS) based on structure, composition, regeneration and deadwood to be mapped across the field site with reasonable certainty. The FCS map closely and consistently corresponded to forest types and stand boundaries, indicating that ALS data offer a feasible approach for forest condition mapping and monitoring to advance forest ecological understanding and improve conservation efforts.

Quantifying 3D vegetation structure in wetlands using differently measured airborne laser scanning data

Mapping and quantifying 3D vegetation structure is essential for assessing and monitoring ecosystem structure and function within wetlands. Airborne Laser Scanning (ALS) is a promising data source for developing indicators of 3D vegetation structure, but derived metrics are often not compared with 3D structural field measurements and the acquisition of ALS data is rarely standardized across different remote sensing surveys. Here, we compare a set of Light Detection And Ranging (LiDAR) metrics derived from ALS datasets with varying characteristics to a standardized set of field measurements of vegetation height, biomass and Leaf Area Index (LAI) across three Hungarian lakes (Lake Balaton, Lake Fertő and Lake Tisza). The ALS datasets differed in whether the recording type was full waveform (FWF) or discrete return, and in their point density (4 pt/m2 and 21 pt/m2). A total of eight LiDAR metrics captured radiometric information as well as descriptors of vegetation cover, height and vertical variability. Multivariate regression models with field-based measurements of vegetation height, biomass or LAI as response variable and LiDAR metrics as predictors showed major differences between ALS recording types, and were affected by differences in spatial resolution, temporal offset and seasonality between field and ALS data acquisition. Vegetation height could be estimated with high to intermediate accuracy (FWF ALS data only: R2 = 0.84; combination of ALS datasets: R2 = 0.67), demonstrating its potential as a robust indicator of 3D vegetation structure across different ALS datasets. In contrast, the estimation of biomass and LAI in these wetlands was sensitive to variation in ALS characteristics and to the discrepancies between field and ALS data in terms of spatial resolution, temporal offset and seasonality (biomass: R2 = 0.20–0.22; LAI: R2 = 0.08–0.30). We recommend the use of FWF ALS data within wetlands because it captures more vegetation structural details in dense reed and marshland vegetation. We further suggest that ecologists and remote sensing scientist should better coordinate the simultaneous and standardized acquisition of field and ALS data for testing the robustness of quantitative descriptors of vegetation cover, height and vertical variability within wetlands. This is important for establishing operational and spatially contiguous ALS-based indicators of 3D ecosystem structure across wetlands.

Processing and analysis of airborne full-waveform laser scanning data for the characterization of forest structure and fuel properties

&lt;p class="Bodytext"&gt;This PhD thesis addresses the development of full-waveform airborne laser scanning (ALS&lt;sub&gt;FW&lt;/sub&gt;) processing and analysis methods to characterize the vertical forest structure, in particular the understory vegetation. In this sense, the influence of several factors such as pulse density, voxel parameters (voxel size and assignation value), scan angle at acquisition, radiometric correction and regression methods is analyzed on the extraction of ALS&lt;sub&gt;FW&lt;/sub&gt; metric values and on the estimate of forest attributes. Additionally, a new software tool to process ALS&lt;sub&gt;FW&lt;/sub&gt; data is presented, which includes new metrics related to understory vegetation. On the other hand, occlusion caused by vegetation in the ALS&lt;sub&gt;FW&lt;/sub&gt;, discrete airborne laser scanning (ALS&lt;sub&gt;D&lt;/sub&gt;) and terrestrial laser scanning (TLS) signal is characterized along the vertical structure. Finally, understory vegetation density is detected and determined by ALS&lt;sub&gt;FW&lt;/sub&gt; data, as well as characterized by using the new proposed metrics.&lt;/p&gt;

A comparative assessment of the vertical distribution of forest components using full-waveform airborne, discrete airborne and discrete terrestrial laser scanning data

Laser scanning has the potential to accurately detect the vertical distribution of forest vegetative components. However, limitations are present and vary according to the system’s platform (i.e., terrestrial or airborne) and recording method (i.e., discrete return or full-waveform). Terrestrial configurations detect close objects (i.e., lower vegetation strata) in more detail while airborne configurations detect a more detailed upper strata, with weak backscattered signals from lower strata. Moreover, discrete lidar systems record single or multiple hits from a given pulse at intercepted features in contrast to full-waveform systems, which register the pulse’s complete backscattered signal providing complete vertical profiles. In this study, we examine for a boreal and a Mediterranean forest with contrasted conifer canopy densities: (i) the characterization of the vertical distribution and signal occlusion from three laser scanning configurations: full-waveform airborne (ALSFW), discrete airborne (ALSD), and discrete terrestrial (TLS); (ii) the comparison in the detection of understory vegetation by ALSFW and ALSD using TLS as reference; and (iii) the use of a methodological procedure based on the Gini index concept to group understory vegetation in density classes from both ALSFW and ALSD configurations. Our results demonstrate, firstly, that signal occlusion can be quantified by the rate of pulse reduction independently for data from all three laser scanning configurations. The ALSD configuration was the most affected by signal occlusion, leading to weak signal returns at the lower strata (z < 4 m) where the rate of pulse reduction was highest as a result of dense canopy covers. Secondly, we demonstrated the capabilities for both airborne laser scanning configurations to detect understory vegetation, albeit significantly more accurately with ALSFW. Lastly, we demonstrated the use of the Gini index as an indicator to determine understory vegetation density classes, particularly for ALSFW data in dense canopy cover. We proceed to explain the limitations in detecting the vertical distribution from different configurations, and indicate that understory vegetation density classes may be successfully assigned with ALSFW in contrasted conifer canopy densities.

A Full-Waveform Airborne Laser Scanning Metric Extraction Tool for Forest Structure Modelling. Do Scan Angle and Radiometric Correction Matter?

In the last decade, full-waveform airborne laser scanning (ALSFW) has proven to be a promising tool for forestry applications. Compared to traditional discrete airborne laser scanning (ALSD), it is capable of registering the complete signal going through the different vertical layers of the vegetation, allowing for a better characterization of the forest structure. However, there is a lack of ALSFW software tools for taking greater advantage of these data. Additionally, most of the existing software tools do not include radiometric correction, which is essential for the use of ALSFW data, since extracted metrics depend on radiometric values. This paper describes and presents a software tool named WoLFeX for clipping, radiometrically correcting, voxelizing the waves, and extracting object-oriented metrics from ALSFW data. Moreover, extracted metrics can be used as input for generating either classification or regression models for forestry, ecology, and fire sciences applications. An example application of WoLFeX was carried out to test the influence of the relative radiometric correction and the acquisition scan angle (1) on the ALSFW metric return waveform energy (RWE) values, and (2) on the estimation of three forest fuel variables (CFL: canopy fuel load, CH: canopy height, and CBH: canopy base height). Results show that radiometric differences in RWE values computed from different scan angle intervals (0°–5° and 15°–20°) were reduced, but not removed, when the relative radiometric correction was applied. Additionally, the estimation of height variables (i.e., CH and CBH) was not strongly influenced by the relative radiometric correction, while the model obtained for CFL improved from R2 = 0.62 up to R2 = 0.79 after applying the correction. These results show the significance of the relative radiometric correction for reducing radiometric differences measured from different scan angles and for modelling some stand-level forest fuel variables.

Classification of Tree Species as Well as Standing Dead Trees Using Triple Wavelength ALS in a Temperate Forest

Knowledge about forest structures, particularly of deadwood, is fundamental for understanding, protecting, and conserving forest biodiversity. While individual tree-based approaches using single wavelength airborne laserscanning (ALS) can successfully distinguish broadleaf and coniferous trees, they still perform multiple tree species classifications with limited accuracy. Moreover, the mapping of standing dead trees is becoming increasingly important for damage calculation after pest infestation or biodiversity assessment. Recent advances in sensor technology have led to the development of new ALS systems that provide up to three different wavelengths. In this study, we present a novel method which classifies three tree species (Norway spruce, European beech, Silver fir), and dead spruce trees with crowns using full waveform ALS data acquired from three different sensors (wavelengths 532 nm, 1064 nm, 1550 nm). The ALS data were acquired in the Bavarian Forest National Park (Germany) under leaf-on conditions with a maximum point density of 200 points/m 2 . To avoid overfitting of the classifier and to find the most prominent features, we embed a forward feature selection method. We tested our classification procedure using 20 sample plots with 586 measured reference trees. Using single wavelength datasets, the highest accuracy achieved was 74% (wavelength = 1064 nm), followed by 69% (wavelength = 1550 nm) and 65% (wavelength = 532 nm). An improvement of 8–17% over single wavelength datasets was achieved when the multi wavelength data were used. Overall, the contribution of the waveform-based features to the classification accuracy was higher than that of the geometric features by approximately 10%. Our results show that the features derived from a multi wavelength ALS point cloud significantly improve the detailed mapping of tree species and standing dead trees.

Relative Radiometric Calibration of Airborne LiDAR Data for Archaeological Applications

Airborne laser scanning (ALS) data can provide more than just a topographic data set for archaeological research. During data collection, laser scanning systems also record radiometric information containing object properties, and thus information about archaeological features. Being aware of the physical model of ALS scanning, the radiometric information can be used to calculate material information of the scanned object. The reflectance of an object or material states the amount of energy it reflects for a specific electromagnetic wavelength. However, the collected radiometric data are affected by several factors that cause dissimilar values to be recorded for the same object. Radiometric calibration of such data minimizes these differences in calculated reflectance values of objects, improving their usability for feature detection and visualization purposes. Previous work dealing with calibration of radiometric data in archaeological research has relied on corresponding in-field measurements to acquire calibration values or has only corrected for a limited number of variables. In this paper, we apply a desk-based approach in which radiometric calibration is conducted through the selection of homogenous areas of interest, without the use of in-field measurements. Together with flight and scan parameters, radiometric calibration allows for the estimation of reflectance values for returns of a single full-waveform ALS data collection flight. The resulting data are then processed into a raster reflectance map that approximates a monochromatic illumination-independent true orthoimage at the wavelength of the laser scanner. We apply this approach to data collected for an archaeological research project in western Sicily and discuss the relative merits of the uses of radiometric data in such locations as well as its wider applicability for present and future archaeological and environmental research. In order to make the approach more accessible, we have developed a freely available tool that allows users to apply the calibration procedure to their own data.

Characterizing understory vegetation in Mediterranean forests using full-waveform airborne laser scanning data

The use of laser scanning acquired from the air, or ground, holds great potential for the assessment of forest structural attributes, beyond conventional forest inventory. The use of full-waveform airborne laser scanning (ALSFW) data allows for the extraction of detailed information in different vertical strata compared to discrete ALS (ALSD). Terrestrial laser scanning (TLS) can register lower vertical strata, such as understory vegetation, without issues of canopy occlusion, however is limited in its acquisition over large areas. In this study we examine the ability of ALSFW to characterize understory vegetation (i.e. maximum and mean height, cover, and volume), verified using TLS point clouds in a Mediterranean forest in Eastern Spain. We developed nine full-waveform metrics to characterize understory vegetation attributes at two different scales (3.75 m square subplots and circular plots with a radius of 15 m); with, and without, application of a height filter to the data. Four understory vegetation attributes were estimated at plot level with high R2 values (mean height: R2 = 0.957, maximum height: R2 = 0.771, cover: R2 = 0.871, and volume: R2 = 0.951). The proportion of explained variance was slightly lower at 3.75 m side cells (mean height: R2 = 0.633, maximum height: R2 = 0.470, cover: R2 = 0.581, and volume R2 = 0.651). These results indicate that Mediterranean understory vegetation can be estimated and accurately mapped over large areas with ALSFW. The future use of these types of predictions includes the estimation of ladder fuels, which drive key fire behavior in these ecosystems.

A supervoxel-based vegetation classification via decomposition and modelling of full-waveform airborne laser scanning data

ABSTRACTVegetation classification is a fundamental task in several applications such as forest management, remote sensing-based crop monitoring, and mitigation of plant diseases, digital prototyping of plants, and plant phenotyping, among others. We propose a novel supervoxel-based methodology to accurately detect vegetation from small-footprint full-waveform airborne laser scanning data in urban and mountainous scenes. Mathematically, the full-waveform decomposition and fitting model based on multiple kernels is presented to generate high-density 3D point clouds and their relevant attributes. The homogeneous supervoxels are then generated by using an enhanced probability density clustering (PDC) algorithm. For each supervoxel, we employ latent Dirichlet allocation to obtain supervoxels features through generalisation of geometric and full-waveform features of point clouds. The Support Vector Machine (SVM) and ensemble classifier random forest (RF) are used to classify these supervoxels into vegetation and non-vegetation. Our experiments on urban and mountainous scenes demonstrate that our approach achieves an overall accuracy of 98.27% and 96.47% respectively by RF classifier and achieves an overall accuracy of 98.16% and 97.67% on the same data sets by SVM classifier. By integrating full-waveform information and more meaningful generalised features, our method outperforms state-of-the-art methods at preserving a trade-off between missing alarm rate and false alarm rate.

Retrieval of higher order statistical moments from full-waveform LiDAR data for tree species classification

Tree inventories, inter alia, need information on the tree species. Previous studies aimed at distinguishing tree species based on three dimensional tree structure metrics derived from airborne laser scanning (ALS) point clouds or based on features from full-waveform (FW) laser scanning data provided by today's sensors. Classifications based on FW features mainly use echo amplitude, pulse energy (hereafter referred to as energy) and width, which are typically retrieved by waveform decomposition, often performed using the symmetric Gaussian distribution function.However, for forested areas, the symmetry of the echo shape is potentially modified by multiple scattering and the distribution of scattering elements (e.g. leaves). We assess the potential of processing full-waveform ALS data such that the third and fourth statistical moments, i.e. the echo skewness and the echo kurtosis, can be retrieved in addition to the amplitude, energy and FWHM. We propose a waveform decomposition approach using the skew normal distribution (SND) function, which enables the modelling of skewed echoes. We investigate the difference of tree-crown aggregated SND derived FW features between seven tree species (969 individual trees) in a temperate mixed forest with the aim of detecting the most descriptive echo features. The such derived FW features are tested for species classification.The results reveal that the largest differences across the tree species are in the mean energy of the first echoes (15 out of 21 species pairs show differences), followed by the mean amplitude of the first echoes and the mean skewness of all echoes originating from a single crown (14 out of 21 species pairs show differences against each other for both features).The differentiation of coniferous and deciduous trees benefits from the features derived from the SND decomposition compared to the use of echo amplitude only (0.39 vs. 0.61 in Cohen's κ). As the classification accuracy of the three dominant tree species within the test site only shows a small increase (0.20 vs. 0.26 in Cohen's κ) by adding FW features, we propose the use of such features in combination with features from multispectral data for this purpose.The SND decomposition is comparable to the Gaussian decomposition regarding the decomposition accuracy (RMSE=4.45 vs. RMSE=3.50) and computational cost. Hence, we propose the default use of the SND decomposition, as the SND is a more flexible function, allowing for the modelling of normally distributed echoes, as well as the fitting of skewed echoes, while no limitations regarding the direction of the skewness are introduced. We attribute the difficulties in the tree species classification to the relatively wide ranges of the crown aggregated features within one species, which for some features results in a considerable overlap of the feature ranges across the species.

Full-Waveform Airborne Laser Scanning in Vegetation Studies—A Review of Point Cloud and Waveform Features for Tree Species Classification

In recent years, small-footprint full-waveform airborne laser scanning has become readily available and established for vegetation studies in the fields of forestry, agriculture and urban studies. Independent of the field of application and the derived final product, each study uses features to classify a target object and to assess its characteristics (e.g., tree species). These laser scanning features describe an observable characteristic of the returned laser signal (e.g., signal amplitude) or a quantity of an object (e.g., height-width ratio of the tree crown). In particular, studies dealing with tree species classification apply a variety of such features as input. However, an extensive overview, categorization and comparison of features from full-waveform airborne laser scanning and how they relate to specific tree species are still missing. This review identifies frequently used full-waveform airborne laser scanning-based point cloud and waveform features for tree species classification and compares the applied features and their characteristics for specific tree species detection. Furthermore, limiting and influencing factors on feature characteristics and tree classification are discussed with respect to vegetation structure, data acquisition and processing.

Aboveground Biomass Estimation of Individual Trees in a Coastal Planted Forest Using Full-Waveform Airborne Laser Scanning Data

The accurate estimation of individual tree level aboveground biomass (AGB) is critical for understanding the carbon cycle, detecting potential biofuels and managing forest ecosystems. In this study, we assessed the capability of the metrics of point clouds, extracted from the full-waveform Airborne Laser Scanning (ALS) data, and of composite waveforms, calculated based on a voxel-based approach, for estimating tree level AGB individually and in combination, over a planted forest in the coastal region of east China. To do so, we investigated the importance of point cloud and waveform metrics for estimating tree-level AGB by all subsets models and relative weight indices. We also assessed the capability of the point cloud and waveform metrics based models and combo model (including the combination of both point cloud and waveform metrics) for tree-level AGB estimation and evaluated the accuracies of these models. The results demonstrated that most of the waveform metrics have relatively low correlation coefficients (&lt;0.60) with other metrics. The combo models (Adjusted R2 = 0.78–0.89), including both point cloud and waveform metrics, have a relatively higher performance than the models fitted by point cloud metrics-only (Adjusted R2 = 0.74–0.86) and waveform metrics-only (Adjusted R2 = 0.72–0.84), with the mostly selected metrics of the 95th percentile height (H95), mean of height of median energy (HOMEμ) and mean of the height/median ratio (HTMRμ). Based on the relative weights (i.e., the percentage of contribution for R2) of the mostly selected metrics for all subsets, the metric of 95th percentile height (H95) has the highest relative importance for AGB estimation (19.23%), followed by 75th percentile height (H75) (18.02%) and coefficient of variation of heights (Hcv) (15.18%) in the point cloud metrics based models. For the waveform metrics based models, the metric of mean of height of median energy (HOMEμ) has the highest relative importance for AGB estimation (17.86%), followed by mean of the height/median ratio (HTMRμ) (16.23%) and standard deviation of height of median energy (HOMEσ) (14.78%). This study demonstrated benefits of using full-waveform ALS data for estimating biomass at tree level, for sustainable forest management and mitigating climate change by planted forest, as China has the largest area of planted forest in the world, and these forests contribute to a large amount of carbon sequestration in terrestrial ecosystems.

Local-scale flood mapping on vegetated floodplains from radiometrically calibrated airborne LiDAR data

Knowledge about the magnitude of localised flooding of riverine areas is crucial for appropriate land management and administration at regional and local levels. However, detection and delineation of localised flooding with remote sensing techniques are often hampered on floodplains by the presence of herbaceous vegetation. To address this problem, this study presents the application of full-waveform airborne laser scanning (ALS) data for detection of floodwater extent. In general, water surfaces are characterised by low values of backscattered energy due to water absorption of the infrared laser shots, but the exact strength of the recorded laser pulse depends on the area covered by the targets located within a laser pulse footprint area. To account for this we analysed the physical quantity of radiometrically calibrated ALS data, the backscattering coefficient, in relation to water and vegetation coverage within a single laser footprint. The results showed that the backscatter was negatively correlated to water coverage, and that of the three distinguished classes of water coverage (low, medium, and high) only the class with the largest extent of water cover (>70%) had relatively distinct characteristics that can be used for classification of water surfaces. Following the laser footprint analysis, three classifiers, namely AdaBoost with Decision Tree, Naïve Bayes and Random Forest, were utilised to classify laser points into flooded and non-flooded classes and to derive the map of flooding extent. The performance of the classifiers is highly dependent on the set of laser points features used. Best performance was achieved by combining radiometric and geometric laser point features. The accuracy of flooding maps based solely on radiometric features resulted in overall accuracies of up to 70% and was limited due to the overlap of the backscattering coefficient values between water and other land cover classes. Our point-based classification methods assure a high mapping accuracy (∼89%) and demonstrate the potential of using full-waveform ALS data to detect water surfaces on floodplain areas with limited water surface exposition through the vegetation canopy.

URBAN TREE CLASSIFICATION USING FULL-WAVEFORM AIRBORNE LASER SCANNING

Vegetation mapping in urban environments plays an important role in biological research and urban management. Airborne laser scanning provides detailed 3D geodata, which allows to classify single trees into different taxa. Until now, research dealing with tree classification focused on forest environments. This study investigates the object-based classification of urban trees at taxonomic family level, using full-waveform airborne laser scanning data captured in the city centre of Vienna (Austria). The data set is characterised by a variety of taxa, including deciduous trees (beeches, mallows, plane trees and soapberries) and the coniferous pine species. A workflow for tree object classification is presented using geometric and radiometric features. The derived features are related to point density, crown shape and radiometric characteristics. For the derivation of crown features, a prior detection of the crown base is performed. The effects of interfering objects (e.g. fences and cars which are typical in urban areas) on the feature characteristics and the subsequent classification accuracy are investigated. The applicability of the features is evaluated by Random Forest classification and exploratory analysis. The most reliable classification is achieved by using the combination of geometric and radiometric features, resulting in 87.5% overall accuracy. By using radiometric features only, a reliable classification with accuracy of 86.3% can be achieved. The influence of interfering objects on feature characteristics is identified, in particular for the radiometric features. The results indicate the potential of using radiometric features in urban tree classification and show its limitations due to anthropogenic influences at the same time.

URBAN TREE CLASSIFICATION USING FULL-WAVEFORM AIRBORNE LASER SCANNING

Abstract. Vegetation mapping in urban environments plays an important role in biological research and urban management. Airborne laser scanning provides detailed 3D geodata, which allows to classify single trees into different taxa. Until now, research dealing with tree classification focused on forest environments. This study investigates the object-based classification of urban trees at taxonomic family level, using full-waveform airborne laser scanning data captured in the city centre of Vienna (Austria). The data set is characterised by a variety of taxa, including deciduous trees (beeches, mallows, plane trees and soapberries) and the coniferous pine species. A workflow for tree object classification is presented using geometric and radiometric features. The derived features are related to point density, crown shape and radiometric characteristics. For the derivation of crown features, a prior detection of the crown base is performed. The effects of interfering objects (e.g. fences and cars which are typical in urban areas) on the feature characteristics and the subsequent classification accuracy are investigated. The applicability of the features is evaluated by Random Forest classification and exploratory analysis. The most reliable classification is achieved by using the combination of geometric and radiometric features, resulting in 87.5% overall accuracy. By using radiometric features only, a reliable classification with accuracy of 86.3% can be achieved. The influence of interfering objects on feature characteristics is identified, in particular for the radiometric features. The results indicate the potential of using radiometric features in urban tree classification and show its limitations due to anthropogenic influences at the same time.

Bottom-up delineation of individual trees from full-waveform airborne laser scans in a structurally complex eucalypt forest

Abstract Full-waveform airborne laser scanning (ALS) is a powerful tool for characterizing and monitoring forest structure over large areas at the individual tree level. Most of the existing ALS-based algorithms for individual tree delineation from the point cloud are top-down, which are accurate for delineating cone-shaped conifers, but have lower delineation accuracies over more structurally complex broad-leaf forests. Therefore, in this study we developed a new bottom-up algorithm for detecting trunks and delineating individual trees with complex shapes, such as eucalypts. Experiments were conducted in the largest river red gum forest in the world, located in the south-east of Australia, that experienced severe dieback over the past six decades. For detection of individual tree trunks, we used a novel approach based on conditional Euclidean distance clustering that takes advantage of spacing between laser returns. Overall, the algorithm developed in our study was able to detect up to 67% of field-measured trees with diameter larger than or equal to 13 cm. By filtering ALS based on the intensity, return number and returned pulse width values, we were able to differentiate between woody and leaf tree components, thus improving the accuracy of tree trunk detections by 5% as compared to non-filtered ALS. The detected trunks were used to seed random walks on graph algorithm for tree crown delineation. The accuracy of tree crown delineation for different ALS point cloud densities was assessed in terms of tree height and crown width and resulted in up to 68% of field-measured trees being correctly delineated. The double increase in point density from ~ 12 points/m2 to ~ 24 points/m2 resulted in tree trunk detection increase of 11% (from 56% to 67%) and percentage of correctly delineated crowns increase of 13% (from 55% to 68%). Our results confirm an algorithm that can be used to accurately delineate individual trees with complex structures (e.g. eucalypts and other broadleaves) and highlight the importance of full-waveform ALS for individual tree delineation.

REFERENCE VALUE PROVISION SCHEMES FOR ATTENUATION CORRECTION OF FULL-WAVEFORM AIRBORNE LASER SCANNER DATA

Abstract. The characterization of the vertical forest structure is highly relevant for ecological research and for better understanding forest ecosystems. Full-waveform airborne laser scanner systems providing a complete time-resolved digitization of every laser pulse echo may deliver very valuable information on the biophysical structure in forest stands. To exploit the great potential offered by full-waveform airborne laser scanning data, the development of suitable voxel based data analysis methods is straightforward. Beyond extracting additional 3D points, it is very promising to derive voxel attributes from the digitized waveform directly. However, the ’history’ of each laser pulse echo is characterized by attenuation effects caused by reflections in higher regions of the crown. As a result, the received waveform signals within the canopy have a lower amplitude than it would be observed for an identical structure without the previous canopy structure interactions (Romanczyk et al., 2012). To achieve a radiometrically correct voxel space representation, the loss of signal strength caused by partial reflections on the path of a laser pulse through the canopy has to be compensated by applying suitable attenuation correction models. The basic idea of the correction procedure is to enhance the waveform intensity values in lower parts of the canopy for portions of the pulse intensity, which have been reflected in higher parts of the canopy. To estimate the enhancement factor an appropriate reference value has to be derived from the data itself. Based on pulse history correction schemes presented in previous publications, the paper will discuss several approaches for reference value estimation. Furthermore, the results of experiments with two different data sets (leaf-on/leaf-off) are presented.

Normalization of Echo Features Derived from Full-Waveform Airborne Laser Scanning Data

Full-waveform airborne laser scanning systems provide fundamental observations for each echo, such as the echo width and amplitude. Geometric and physical information about illuminated surfaces are simultaneously provided by a single scanner. However, there are concerns about whether the physical meaning of observations is consistent among different scanning missions. Prior to the application of waveform features for multi-temporal data classification, such features must be normalized. This study investigates the transferability of normalized waveform features to different surveys. The backscatter coefficient is considered to be a normalized physical feature. A normalization process for the echo width, which is a geometric feature, is proposed. The process is based on the coefficient of variation of the echo widths in a defined neighborhood, for which the Fuzzy Small membership function is applied. The normalized features over various land cover types and flight missions are investigated. The effects of different feature combinations on the classification accuracy are analyzed. The overall accuracy of the combination of normalized features and height-based attributes achieves promising results (&gt;93% overall accuracy for ground, roof, low vegetation, and tree canopy) when different flight missions and classifiers are used. Nevertheless, the combination of all possible features, including raw features, normalized features, and height-based features, performs less well and yields inconsistent results.

Correcting attenuation effects caused by interactions in the forest canopy in full-waveform airborne laser scanner data

Abstract. Full-waveform airborne laser scanning offers a great potential for various forestry applications. Especially applications requiring information on the vertical structure of the lower canopy parts benefit from the great amount of information contained in waveform data. To enable the derivation of vertical forest canopy structure, the development of suitable voxel based data analysis methods is straightforward. Beyond extracting additional 3D points, it is very promising to derive the voxel attributes from the digitized waveform directly. For this purpose, the differential backscatter cross sections have to be projected into a Cartesian voxel structure. Thereby the voxel entries represent amplitudes of the cross section and can be interpreted as a local measure for the amount of pulse reflecting matter. However, the "history" of each laser echo pulse is characterized by attenuation effects caused by reflections in higher regions of the crown. As a result, the received waveform signals within the canopy have a lower amplitude than it would be observed for an identical structure without the previous canopy structure interactions (Romanczyk et al., 2012). If the biophysical structure is determined from the raw waveform data, material in the lower parts of the canopy is thus under-represented. To achieve a radiometrically correct voxel space representation the loss of signal strength caused by partial reflections on the path of a laser pulse through the canopy has to be compensated. In this paper, we present an integral approach correcting the waveform at each recorded sample. The basic idea of the procedure is to enhance the waveform intensity values in lower parts of the canopy for portions of the pulse intensity, which have been reflected (and thus blocked) in higher parts of the canopy. The paper will discuss the developed correction method and show results from a validation both with synthetic and real world data.