Raw Light Detection And Ranging Research Articles

Automatic detection of single street trees from airborne LiDAR data based on point segmentation methods

As a primary element of urban ecosystem, street trees are very essential for environmental quality and aesthetic beauty of urban landscape. Street trees play a crucial role in everyday life of city inhabitants and therefore, comprehensive and accurate inventory information for street trees is required. In this research, an automatic method is proposed to detect single street trees from airborne Light Detection and Ranging (LiDAR) point cloud instead of traditional field work or photo interpretation. Firstly, raw LiDAR point cloud data have been classified to obtain high vegetation class with a hierarchical rule-based classification method. Then, the LiDAR points in high vegetation class were segmented with mean shift and Density Based Spatial Clustering of Applications with Noise (DBSCAN) algorithms to acquire single urban street trees in the Davutpasa Campus of Yildiz Technical University, Istanbul, Turkey. The accuracy assessment of the acquired street trees was also conducted using completeness and correctness analyses. The acquired results from urban study area approved the success of the proposed point-based approach for automatic detection of single street trees using LiDAR point cloud.

An Efficient Adaptive Noise Removal Filter on Range Images for LiDAR Point Clouds

Light Detection and Ranging (LiDAR) is a critical sensor for autonomous vehicle systems, providing high-resolution distance measurements in real-time. However, adverse weather conditions such as snow, rain, fog, and sun glare can affect LiDAR performance, requiring data preprocessing. This paper proposes a novel approach, the Adaptive Outlier Removal filter on range Image (AORI), which combines a projection image from LiDAR point clouds with an adaptive outlier removal filter to remove snow particles. Our research aims to analyze the characteristics of LiDAR and propose an image-based approach derived from LiDAR data that addresses the limitations of previous studies, particularly in improving the efficiency of nearest neighbor point search. Our proposed method achieves outstanding performance in both accuracy (>96%) and processing speed (0.26 s per frame) for autonomous driving systems under harsh weather from raw LiDAR point clouds in the Winter Adverse Driving dataset (WADS). Notably, AORI outperforms state-of-the-art filters by achieving a 6.6% higher F1 score and 0.7% higher accuracy. Although our method has a lower recall than state-of-the-art methods, it achieves a good balance between retaining object points and filter noise points from LiDAR, indicating its promise for snow removal in adverse weather conditions.

AFLI-Calib: Robust LiDAR-IMU extrinsic self-calibration based on adaptive frame length LiDAR odometry

As an effective complement to common laser scanning systems, the portable laser scanning system can acquire point clouds flexibly and quickly. Calibration between Light detection and ranging (LiDAR) sensors and inertial measurement units (IMU) is the prerequisite for laser scanning systems to obtain high-quality point clouds. Related methods have been proposed in the last two decades, where the global navigation satellite system (GNSS) or high-precision calibration fields are commonly used. However, the extrinsic self-calibration of LiDAR-IMU is challenging, due to the large distortion in single-frame point cloud caused by rapid motion and the position errors of IMU integration which drift quickly. At the same time, the highly dynamic motion patterns of portable devices and the changes in the scanned scene structure are not well considered in existing LiDAR odometry methods. To take better advantage of the characteristics of non-repetitive scanning LiDAR sensor, this paper proposes AFLI-Calib, which utilizes adaptive frame length LiDAR odometry to perform the extrinsic self-calibration of LiDAR-IMU. Unlike LiDAR odometry methods with a fixed frame length, the LiDAR frame length is dynamically adjusted according to the motion state of sensors and the matching stability of scenes. The single-frame point cloud is registered to the map through a linear-based continuous-time model, eliminating the motion distortion correction in advance. For further optimization of trajectory and extrinsic parameters, IMU raw measurements and LiDAR observations are involved in the multi-constraint optimization, through tightly-coupled IMU pre-integration constraints, LiDAR point-to-plane constraints, and prior constraints. The method is fully validated using self-collected calibration data of indoor and outdoor scenes and different motion modes. Experiments show that on the test data, the translation parameter accuracy of the method is 0.041 m, which is 56.3% higher than the state-of-the-art method. The standard deviation is significantly reduced, with translation deviation (0.017 m, 0.024 m, 0.022 m) and rotation deviation (0.17°, 0.25°, 0.15°), which verifies the robustness of our method. The average RMSE of distances to the reference point cloud acquired by the terrestrial laser scanning system (TLS) is 0.042 m, showing a high accuracy calibration result. Comparative experiments with the fixed frame length LiDAR odometry method and classical “correction-then-registration” motion distortion model further verify the superiority and effectiveness of the proposed adaptive frame length LiDAR odometry.

An Adaptive Group of Density Outlier Removal Filter: Snow Particle Removal from LiDAR Data

Light Detection And Ranging (LiDAR) is an important technology integrated into self-driving cars to enhance the reliability of these systems. Even with some advantages over cameras, it is still limited under extreme weather conditions such as heavy rain, fog, or snow. Traditional methods such as Radius Outlier Removal (ROR) and Statistical Outlier Removal (SOR) are limited in their ability to detect snow points in LiDAR point clouds. This paper proposes an Adaptive Group of Density Outlier Removal (AGDOR) filter that can remove snow particles more effectively in raw LiDAR point clouds, with verification on the Winter Adverse Driving Dataset (WADS). In our proposed method, an intensity threshold combined with a proposed outlier removal filter was employed. Outstanding performance was obtained, with higher accuracy up to 96% and processing speed of 0.51 s per frame in our result. In particular, our filter outperforms the state-of-the-art filter by achieving a 16.32% higher Precision at the same accuracy. However, our method archive is lower in recall than the state-of-the-art method. This clearly indicates that AGDOR retains a significant amount of object points from LiDAR. The results suggest that our filter would be useful for snow removal under harsh weathers for autonomous driving systems.

Potential and limitations of LiDAR altimetry in archaeological survey. Copper Age and Bronze Age settlements in southern Iberia

Potential and limitations of LiDAR altimetry in archaeological survey. Copper Age and Bronze Age settlements in southern Iberia

The Use of Machine Learning Algorithms in Urban Tree Species Classification

Trees are the key components of urban vegetation in cities. The timely and accurate identification of existing urban tree species with their location is the most important task for improving air, water, and land quality; reducing carbon accumulation; mitigating urban heat island effects; and protecting soil and water balance. Light detection and ranging (LiDAR) is frequently used for extracting high-resolution structural information regarding tree objects. LiDAR systems are a cost-effective alternative to the traditional ways of identifying tree species, such as field surveys and aerial photograph interpretation. The aim of this work was to assess the usage of machine learning algorithms for classifying the deciduous (broadleaf) and coniferous tree species from 3D raw LiDAR data on the Davutpasa Campus of Yildiz Technical University, Istanbul, Turkey. First, ground, building, and low, medium, and high vegetation classes were acquired from raw LiDAR data using a hierarchical-rule-based classification method. Next, individual tree crowns were segmented using a mean shift clustering algorithm from high vegetation points. A total of 25 spatial- and intensity-based features were utilized for support vector machine (SVM), random forest (RF), and multi-layer perceptron (MLP) classifiers to discriminate deciduous and coniferous tree species in the urban area. The machine learning-based classification’s overall accuracies were 80%, 83.75%, and 73.75% for the SVM, RF, and MLP classifiers, respectively, in split 70/30 (training/testing). The SVM and RF algorithms generally gave better classification results than the MLP algorithm for identifying the urban tree species.

Classification of raw LiDAR point cloud using point-based methods with spatial features for 3D building reconstruction

Building extraction from light detection and ranging (LiDAR) data for 3-dimensional (3D) reconstruction requires accurately classified LiDAR points. In recent years, approaches developed for the classification mostly based on gridded LiDAR data. In the gridding process of LiDAR data, there is a characteristic point loss which results in reduced height accuracy. The effect of such loss can be eliminated using classified raw LiDAR data. In this study, an automatic point-based classification approach for raw LiDAR data classification with spatial features has been proposed for 3D building reconstruction. Using spatial features, the hierarchical rules have been determined. The spatial features, such as height, the local environment, and multi-return, of the LiDAR points were analyzed, and every single LiDAR points automatically assigned to the classes based on these features. The proposed classification approach based on raw LiDAR data had an overall accuracy of 79.7% in the test site located in Istanbul, Turkey. Finally, 3D building reconstruction was performed using the results of the proposed automatic point-based classification approach.

Estimating above-ground biomass of subtropical forest using airborne LiDAR in Hong Kong

Seventy-percent of the terrestrial area of Hong Kong is covered by vegetation and 40% is protected as the Country Park. The above-ground biomass (AGB) acts as reliable source of carbon sink and while Hong Kong has recognized the importance of carbon sink in forest and urged for forest protection in the latest strategic plan, yet no study has been conducted on assessing the baseline of terrestrial AGB and its carbon storage. This study compared and estimated the AGB by the traditional allometric modeling and the Light Detection and Ranging (LiDAR) plot metrics at plot-level in a subtropical forest of Hong Kong. The study has tested five allometric models which were developed from pantropical regions, subtropical areas and locally. The best model was then selected as the dependent variable to develop the LiDAR-derived AGB model. The raw LiDAR point cloud was pre-processed to normalized height point cloud and hence generating the LiDAR metric as independent variables for the model development. Regression models were used to estimate AGB at various plot sizes (i.e., in 10-m, 5-m and 2.5-m radius). The models were then evaluated statistically and validated by bootstrapping and leave-one-out cross validation (LOOCV). The results indicated the LiDAR metric derived from larger plot size outperformed the smaller plot size, with model R2 of 0.864 and root-mean-square-error (RMSE) of 37.75 kg/ha. It also found that pantropical model was comparable to a site-specific model when including the bioclimatic variable in subtropical forests. This study provides the approach for delineating the baseline of terrestrial above-ground biomass and carbon stock in subtropical forests upon an appropriate plot size is being deployed.

LiDAR-INS/GNSS-Based Real-Time Ground Removal, Segmentation, and Georeferencing Framework for Smart Transportation

Light detection and ranging (LiDAR) sensor is attracting significant attention in the field of smart transportation because of its ability to give depth information. It is already extensively used for obstacle detection. The inertial navigation system (INS) with a global navigation satellite system (GNSS) gives the global position, orientation, and velocity of a given object in real time. In autonomous driving, LiDAR point clouds are segmented to find out the positions of various objects in the surrounding of the vehicle of interest. However, the obstacle position is determined in local east–north–up (ENU) coordinates. In this article, an end-to-end framework has been presented, which takes input from LiDAR and INS/GNSS systems and gives the obstacle position in universal coordinates (latitude, longitude) in real time. The proposed framework includes ground point removal from raw LiDAR data, obstacle segmentation, and LiDAR data fusion with INS/GNSS data for georeferencing. For LiDAR data processing, a novel method for removing ground points has been presented, in which the entire point cloud is divided into square grids in the horizontal plane. Based on the statistics of the vertical distribution of the points in each grid, ground points are identified. Two ground point datasets were also created for testing the proposed algorithm. These datasets contain point clouds of ground with varying inclination from various multichannel LiDARs. The proposed ground removal method was tested on the Paris-Lille-3D dataset and the dataset created. An F1-score of greater than 0.99 and 0.98 was achieved on our dataset and the Paris Lille-3D dataset, respectively. The framework is tested on different hardware configurations and found to be suitable for real-time applications.

Automatic Weighted Splines Filter (AWSF): A New Algorithm for Extracting Terrain Measurements From Raw LiDAR Point Clouds

Airborne raw light detection and ranging (LiDAR) measurements are georeferenced three-dimensional coordinates of ground surface, including all natural and man-made features. Extracting terrain surface measurements from raw LiDAR data is referred to as “filtering.” Many filtering algorithms have been published, indicating the difficulty of the task; however, none performs equally well on all kinds of landscapes. This article presents a new algorithm, automatic weighted splines filter (AWSF), to extract the terrain points from raw LiDAR measurements. The mathematical model of the AWSF algorithm utilizes both the cubic smoothing splines (to interpolate and fit raw LiDAR data) and z-shaped function (to estimate the weight value of each point). The AWSF algorithm performance is compared against 14 filtering algorithms published between 1998 and 2019, as well as one unpublished (proprietary) algorithm designed by the world-leading company in processing remote sensing data, Harris Geospatial Solutions. Diverse landscape scenarios are used, ranging from open fields, rural land, and urban areas to dense forests on mountains where most of the filtering algorithms struggle as these areas represent the most difficult and challenging landscapes because the canopy prevents LiDAR pulses from reaching the ground surface. A total of 19 samples were tested; the results clearly show that the filtered terrain measurements are accurate and that the performance of the AWSF algorithm is stable for all the LiDAR data samples in comparison with the other filtering approaches.

A single-shot pose estimation approach for a 2D laser rangefinder

Two-dimensional (2D) laser rangefinders (LRFs) are widely utilized in various mobilized systems, such as robots, intelligent vehicles (IVs) and mobile mapping. Pose estimation for a 2D LRF has important applications for exterior calibration of multiple sensors, registering light detection and ranging (LIDAR) data from multiple LRFs or from a spinning/nodding LRF, etc. This paper shows that the pose of a 2D LRF can be uniquely determined from a single-shot of a trirectangular trihedron. The pose is estimated by solving a simplified perspective-three-point (P3P) problem and a three-point-registration problem. Compared to the raw noisy LIDAR data, the three control points in the P3P problem are estimated from line fitting of the sensed LIDAR data to guarantee good and reliable estimation results. The proposed method has been validated with both simulation and real data. The results show that the method is accurate and practical. To the best of our knowledge, this paper solves the pose estimation of a 2D LRF by using LIDAR data from one scan for the first time.

Building Extraction From LiDAR Data Applying Deep Convolutional Neural Networks

Deep learning paradigm has been shown to be a very efficient classification framework for many application scenarios, including the analysis of Light Detection and Ranging (LiDAR) data for building detection. In fact, deep learning acts as a set of mathematical transformations, encoding the raw input data into appropriate forms of representations that maximize the classification performance. However, it is clear that mathematical computations alone, even highly nonlinear, are not adequate to model the physical properties of a problem, distinguishing, for example, the building structures from vegetation. In this letter, we address this difficulty by augmenting the raw LiDAR data with features coming from a physical interpretation of the information. Then, we exploit a deep learning paradigm based on a convolutional neural network model to find out the best input representations suitable for the classification. As test sites, three complex urban study areas with various kinds of building structures through the LiDAR data set of Vaihingen, Germany were selected. Our method has been evaluated in the context of “ISPRS Test Project on Urban Classification and 3-D Building Reconstruction.” Comparisons with traditional methods, such as artificial neural networks and support vector machine-based classifiers, indicate the outperformance of the proposed approach in terms of robustness and efficiency.

A Density-Based Clustering Method for the Segmentation of Individual Buildings from Filtered Airborne LiDAR Point Clouds

Individual building segmentation is a prerequisite for building reconstruction. When building points or building regions are classified from raw LiDAR (Light Detection and Ranging) point clouds, the dataset usually contains numerous individual buildings as well as outliers. However, the applications to segment individual buildings from large datasets require the algorithms working with the minimal requirements of domain knowledge to determine the input parameters, working well on datasets with outliers and having good efficiency on big data. To meet these requirements, this paper presents a new segmentation method relying on a density-based clustering technique that is designed to separate individual buildings in dense built-up areas and is robust to outliers. As implemented in a spatial database, the algorithm benefits from the spatial index and the parallel computation capability offered by the system. The experimental results show that the proposed method is significantly more effective in segmenting individual buildings than the well-known moving window algorithm and the new boundary identification and tracing algorithm, and processes large volumes of data with good efficiency. Compared with the moving window algorithm, the proposed method (parallelized) consumed only 17.8% time and the quality improved from 88.8 to 94.8% on the Vaihingen dataset.

Generation of digital terrain model for forest areas using a new particle swarm optimization on LiDAR data

Since Light Detection and Ranging (LiDAR) data are capable of distinguishing vegetation from bare earth, these data are used nowadays to produce digital terrain models (DTMs) for forest regions. In this research, raw LiDAR data were filtered using hybrid and slope-based filtering methods and the filtered data were then interpolated using the new modified particle swarm optimisation (PSO) and accordingly the results were compared with those achieved by the other intelligent and conventional interpolation methods. The new modified PSO optimized the polynomial degree for interpolation and found suitable parameters for optimisation. Two data sets from two forest regions in some northern regions of Iran located in Golestan province were selected to compare these methods. Region 1 with dense vegetation and region 2 with grass vegetation. The results indicated that the hybrid filter performed lower RMSE than the slope-based filter. Finally, the DTM with lowest RMSE was obtained using the hybrid filter and the modified PSO interpolation method with RMSE of 6 mm for region 1 (Tavar-kuh) and 61 mm for region 2 (Shastkola River Basin).

INDIVIDUAL TREE CROWN DELINEATION USING MULTI-WAVELENGTH TITAN LIDAR DATA

Abstract. The inability to detect the Emerald Ash Borer (EAB) at an early stage has led to the enumerable loss of different species of ash trees. Due to the increasing risk being posed by the EAB, a robust and accurate method is needed for identifying Individual Tree Crowns (ITCs) that are at a risk of being infected or are already diseased. This paper attempts to outline an ITC delineation method that employs airborne multi-spectral Light Detection and Ranging (LiDAR) to accurately delineate tree crowns. The raw LiDAR data were initially pre-processed to generate the Digital Surface Models (DSM) and Digital Elevation Models (DEM) using an iterative progressive TIN (Triangulated Irregular Network) densification method. The DSM and DEM were consequently used for Canopy Height Model (CHM) generation, from which the structural information pertaining to the size and shape of the tree crowns was obtained. The structural information along with the spectral information was used to segment ITCs using a region growing algorithm. The availability of the multi-spectral LiDAR data allows for delineation of crowns that have otherwise homogenous structural characteristics and hence cannot be isolated from the CHM alone. This study exploits the spectral data to derive initial approximations of individual tree tops and consequently grow those regions based on the spectral constraints of the individual trees.

Comparing the capability of low- and high-resolution LiDAR data with application to solar resource assessment, roof type classification and shading analysis

LiDAR (Light Detection and Ranging) data have recently gained popularity for use in solar resource assessment and solar photovoltaics (PV) suitability studies in the built environment due to robustness at identifying building orientation, roof tilt and shading. There is a disparity in the geographic coverage of low- and high-resolution LiDAR data (LL and LH, respectively) between rural and urban locations, as the cost of the latter is often not justified for rural areas where high PV penetrations often pose the greatest impact on the electricity distribution network. There is a need for a comparison of the different resolutions to assess capability of LL. In this study, we evaluated and improved upon a previously reported methodology that derives roof types from a LiDAR-derived, low-resolution Digital Surface Model (DSM) with a co-classing routine. Key improvements to the methodology include: co-classing routine adapted for raw LiDAR data, applicability to differing building type distribution in study area, building height and symmetry considerations, a vector-based shading analysis of building surfaces and the addition of solar resource assessment capability.Based on the performance of different LiDAR resolutions within the developed model, a comparison between LL (0.5–1pts/m2) and LH (6–8pts/m2) LiDAR data was applied; LH can confidently be used to evaluate the applicability of LL, due to its significantly higher point density and therefore accuracy. We find that the co-classing methodology works satisfactory for LL for all types of building distributions. Roof-type identification errors from incorrect co-classing were rare (<1%) with LL. Co-classing buildings using LL improves accuracy of roof-type identification in areas with homogeneous distribution of buildings, here from 78% to 86% in accuracy. Contrastingly, co-classing accuracy using LH is marginally reduced for all building distributions from 94.8% to 94.4%. We adapt the Hay and Davies solar transposition model to include shading. The shading analysis demonstrates similarity of results between LL and LH. We find that the proposed methodology can confidently be used for solar resource assessments on buildings when only LiDAR data of low-resolution (<1pts/m2) is available.

Improved progressive TIN densification filtering algorithm for airborne LiDAR data in forested areas

Filtering of light detection and ranging (LiDAR) data into the ground and non-ground points is a fundamental step in processing raw airborne LiDAR data. This paper proposes an improved progressive triangulated irregular network (TIN) densification (IPTD) filtering algorithm that can cope with a variety of forested landscapes, particularly both topographically and environmentally complex regions. The IPTD filtering algorithm consists of three steps: (1) acquiring potential ground seed points using the morphological method; (2) obtaining accurate ground seed points; and (3) building a TIN-based model and iteratively densifying TIN. The IPTD filtering algorithm was tested in 15 forested sites with various terrains (i.e., elevation and slope) and vegetation conditions (i.e., canopy cover and tree height), and was compared with seven other commonly used filtering algorithms (including morphology-based, slope-based, and interpolation-based filtering algorithms). Results show that the IPTD achieves the highest filtering accuracy for nine of the 15 sites. In general, it outperforms the other filtering algorithms, yielding the lowest average total error of 3.15% and the highest average kappa coefficient of 89.53%.

An investigation of DEM generation process based on LiDAR data filtering, decimation, and interpolation methods for an urban area

Terrain analysis is an essential step to assess landscape, spatial variability of soil and ecologic properties. An easy and effective way of this is studying with Digital Elevation Model (DEM). Traditional methods for creating DEM are very costly and time consuming because of land surveying. In time, Photogrammetry has become one of the major methods to generate DEM. Recently, airborne Light Detection and Ranging (LiDAR) system has become a powerful way to produce a DEM due to advantage of collecting three-dimensional information very effectively over a large area by means of precision and time. Airborne LiDAR system collects information not only from land surface, but also from every object between plane and terrain that can reflect the laser beam. Therefore, LiDAR data density, topographic features, and interpolation methods are the main parameters for LiDAR derived DEM accuracy.In this paper, an investigation is carried out for generating a proper DEM to analyze terrain futures. For this purpose, the DEM generation process is started with raw LiDAR data of an urban area filtering with Triangular Irregular Network (TIN) algorithm before and after for five different decimation rates. Afterward, the interpolation techniques that are the Inverse Distance Weighted (IDW), Kriging, Natural Neighbor (NN), and Topo to Raster, are used for DEM generation. According to the accuracy assessments, the TIN algorithm is very satisfactory for filtering process and the NN interpolation method is quite effective for DEM generation.

Automatic Segmentation of Raw LIDAR Data for Extraction of Building Roofs

Automatic extraction of building roofs from remote sensing data is important for many applications, including 3D city modeling. This paper proposes a new method for automatic segmentation of raw LIDAR (light detection and ranging) data. Using the ground height from a DEM (digital elevation model), the raw LIDAR points are separated into two groups. The first group contains the ground points that form a “building mask”. The second group contains non-ground points that are clustered using the building mask. A cluster of points usually represents an individual building or tree. During segmentation, the planar roof segments are extracted from each cluster of points and refined using rules, such as the coplanarity of points and their locality. Planes on trees are removed using information, such as area and point height difference. Experimental results on nine areas of six different data sets show that the proposed method can successfully remove vegetation and, so, offers a high success rate for building detection (about 90% correctness and completeness) and roof plane extraction (about 80% correctness and completeness), when LIDAR point density is as low as four points/m2. Thus, the proposed method can be exploited in various applications.

Automatic extraction of building roofs using LIDAR data and multispectral imagery

Automatic 3D extraction of building roofs from remotely sensed data is important for many applications including city modelling. This paper proposes a new method for automatic 3D roof extraction through an effective integration of LIDAR (Light Detection And Ranging) data and multispectral orthoimagery. Using the ground height from a DEM (Digital Elevation Model), the raw LIDAR points are separated into two groups. The first group contains the ground points that are exploited to constitute a ‘ground mask’. The second group contains the non-ground points which are segmented using an innovative image line guided segmentation technique to extract the roof planes. The image lines are extracted from the grey-scale version of the orthoimage and then classified into several classes such as ‘ground’, ‘tree’, ‘roof edge’ and ‘roof ridge’ using the ground mask and colour and texture information from the orthoimagery. During segmentation of the non-ground LIDAR points, the lines from the latter two classes are used as baselines to locate the nearby LIDAR points of the neighbouring planes. For each plane a robust seed region is thereby defined using the nearby non-ground LIDAR points of a baseline and this region is iteratively grown to extract the complete roof plane. Finally, a newly proposed rule-based procedure is applied to remove planes constructed on trees. Experimental results show that the proposed method can successfully remove vegetation and so offers high extraction rates.