Digital Terrain Model Generation Research Articles

Generation of a Digital Terrain Model (DTM) Fusioning WV-2 Images and RTK-derived Topobathymetric Data

Digital terrain models (DTMs) are digital elevation models (DEMs) that represent the bare ground surface. They are created by multiple sources, including satellite remote sensing, aerial photography, and ground-based surveys, and are often combined with other data sources to create highly detailed models. As the demand for accurate and detailed information about the Earth's surface continues to grow, DTMs have become an increasingly important tool for researchers in different fields. This study aims to create a DTM with a spatial resolution of 0.50 m for São Caetano do Sul, São Paulo, Brazil, integrated with a topobathymetric map of three water courses running along the borders of the study area. For the conventional DTM generation, a WV-2 stereo pair was used. A total of 55 ground control points (GCPs) were collected using the GNSS-RTK method, being 60% used for model building and 40% employed for validation. The topobathymetric survey was accomplished using a GNSS-RTK device placed along the analyzed open streams. For validation purposes, we used bias and MAE metrics. Overall, the methodology presented in this article provides a useful approach for generating high-resolution DTMs that can be used in a range of applications, especially in urban hydrodynamic studies.

Digital terrain, surface, and canopy height model generation with dual-baseline low-frequency InSAR over forest areas

Digital terrain, surface, and canopy height model generation with dual-baseline low-frequency InSAR over forest areas

Stereophotoclinometry for OSIRIS-REx Spacecraft Navigation

We summarize a decade of effort by the Origins, Spectral Interpretation, Resource Identification, and Security–Regolith Explorer (OSIRIS-REx) mission team to build up the unique capabilities, processes, and procedures required to accomplish the unprecedented navigation performance required during proximity operations at asteroid (101955) Bennu. Stereophotoclinometry was a key technology used for digital terrain model (DTM) generation and landmark navigation, enabling estimation of spacecraft trajectories and Bennu’s geophysical parameters. We outline the concept of operations for OSIRIS-REx landmark navigation and the wide array of testing and verification efforts leading up to OSIRIS-REx’s arrival at Bennu. We relate the outcome of these efforts to the experiences during proximity operations. We discuss navigation and DTM performance during operations, including detailed lessons learned to carry forward for future missions.

An Object-Based Ground Filtering of Airborne LiDAR Data for Large-Area DTM Generation

Digital terrain model (DTM) creation is a modeling process that represents the Earth’s surface. An aptly designed DTM generation method tailored for intended study can significantly streamline ensuing processes and assist in managing errors and uncertainties, particularly in large-area projects. However, existing methods often exhibit inconsistent and inexplicable results, struggle to clearly define what an object is, and often fail to filter large objects due to their locally confined operations. We introduce a new DTM generation method that performs object-based ground filtering, which is particularly beneficial for urban topography. This method defines objects as areas fully enclosed by steep slopes and grounds as smoothly connected areas, enabling reliable “object-based” segmentation and filtering, extending beyond the local context. Our primary operation, controlled by a slope threshold parameter, simplifies tuning and ensures predictable results, thereby reducing uncertainties in large-area modeling. Uniquely, our method considers surface water bodies in modeling and treats connected artificial terrains (e.g., overpasses) as ground. This contrasts with conventional methods, which often create noise near water bodies and behave inconsistently around overpasses and bridges, making our approach particularly beneficial for large-area 3D urban mapping. Examined on extensive and diverse datasets, our method offers unique features and high accuracy, and we have thoroughly assessed potential artifacts to guide potential users.

OPTIMUM NUMBER AND POSITIONS OF TERRESTRIAL LASER SCANNER TO DERIVE DTM AT FOREST PLOT LEVEL

Abstract. Remote sensing of forest landscape is sometimes quite challenging, considering the rugged terrain, remote areas and presence of dense canopy. LiDAR is capable of generating dense three-dimensional (3D) structure of the forest. But when research is focused on a small area that needs dense information Terrestrial Laser Scanner (TLS) proves to be an efficient way of acquiring the information. TLS is an instrument used to acquire data from the ground surface level. Ground information and under canopy structure are acquired with more accuracy and precession with TLS when compared to aerial remote sensing techniques. When a TLS survey is done, scan positions and the number of scans are important parameters to consider to acquire datasets of forest plots. In this research, an analysis is carried out to determine the optimum number of TLS scans and positions that are needed to sufficiently generate the number of ground points to generate a Digital Terrain Model (DTM). The research considers two forest plots of 25m × 25m with different tree densities. Different TLS scanning combinations are used to determine the best scanning pattern for DTM generation at the forest plot level.

Practical Stereophotoclinometry for Modeling Shape and Topography on Planetary Missions

Stereophotoclinometry (SPC) is a technique to extract topographic information from images acquired by spacecraft. It combines stereophotogrammetry and photoclinometry to produce a product that has the accuracy of stereo with the resolution of photoclinometry without the restrictions common to both. We describe the implementation of this technique in the context of digital terrain model (DTM) generation for a small-body mission. We detail the process and the data used to generate SPC-derived DTMs at progressively increasing resolutions. The highest-quality DTMs are generated using four images optimized for topography, a 30° emission angle with the emission azimuth (spacecraft position) to the north, east, south, and west of the target, and one image optimized for albedo (a low incidence angle such that most of the image pixels’ digital numbers are based upon albedo rather than topography). We discuss implications for mission planning and how SPC-based DTM generation can support spacecraft navigation. As a case study, we share outcomes from the modeling performed for the OSIRIS-REx mission to asteroid Bennu.

Comparative Analysis of Multi-Platform, Multi-Resolution, Multi-Temporal LiDAR Data for Forest Inventory

LiDAR technology is rapidly evolving as various new systems emerge, providing unprecedented data to characterize forest vertical structure. Data from different LiDAR systems present distinct characteristics owing to a combined effect of sensor specifications, data acquisition strategies, as well as forest conditions such as tree density and canopy cover. Comparative analysis of multi-platform, multi-resolution, and multi-temporal LiDAR data provides guidelines for selecting appropriate LiDAR systems and data processing tools for different research questions, and thus is of crucial importance. This study presents a comprehensive comparison of point clouds from four systems, linear and Geiger-mode LiDAR from manned aircraft and multi-beam LiDAR on unmanned aerial vehicle (UAV), and in-house developed Backpack, with the consideration of different forest canopy cover scenarios. The results suggest that the proximal Backpack LiDAR can provide the finest level of information, followed by UAV LiDAR, Geiger-mode LiDAR, and linear LiDAR. The emerging Geiger-mode LiDAR can capture a significantly higher level of detail while operating at a higher altitude as compared to the traditional linear LiDAR. The results also show: (1) canopy cover percentage has a critical impact on the ability of aerial and terrestrial systems to acquire information corresponding to the lower and upper portions of the tree canopy, respectively; (2) all the systems can obtain adequate ground points for digital terrain model generation irrespective of canopy cover conditions; and (3) point clouds from different systems are in agreement within a ±3 cm and ±7 cm range along the vertical and planimetric directions, respectively.

Simplifying UAV-Based Photogrammetry in Forestry: How to Generate Accurate Digital Terrain Model and Assess Flight Mission Settings

In forestry, aerial photogrammetry by means of Unmanned Aerial Systems (UAS) could bridge the gap between detailed fieldwork and broad-range satellite imagery-based analysis. However, optical sensors are only poorly capable of penetrating the tree canopy, causing raw image-based point clouds unable to reliably collect and classify ground points in woodlands, which is essential for further data processing. In this work, we propose a novel method to overcome this issue and generate accurate a Digital Terrain Model (DTM) in forested environments by processing the point cloud. We also developed a highly realistic custom simulator that allows controlled experimentation with repeatability guaranteed. With this tool, we performed an exhaustive evaluation of the survey and sensor settings and their impact on the 3D reconstruction. Overall, we found that a high frontal overlap (95%), a nadir camera angle (90°), and low flight altitudes (less than 100 m) results in the best configuration for forest environments. We validated the presented method for DTM generation in a simulated and real-world survey missions with both fixed-wing and multicopter UAS, showing how the problem of structural forest parameters estimation can be better addressed. Finally, we applied our method for automatic detection of selective logging.

Digital Deforestation: Comparing Automated Approaches to the Production of Digital Terrain Models (DTMs) in Agisoft Metashape

This paper tests the suitability of automated point cloud classification tools provided by the popular image-based modeling (IBM) software package Agisoft Metashape for the generation of digital terrain models (DTMs) at moderately-vegetated archaeological sites. DTMs are often required for various forms of archaeological mapping and analysis. The suite of tools provided by Agisoft are relatively user-friendly as compared to many point cloud classification algorithms and do not require the use of additional software. Based on a case study from the Mycenaean site of Kastrouli, Greece, the mostly-automated, geometric classification tool “Classify Ground Points” provides the best results and produces a quality DTM that is sufficient for mapping and analysis. Each of the methods tested in this paper can likely be improved through manual editing of point cloud classification.

Assessing spatio-temporal patterns of sugarcane aphid (Hemiptera: Aphididae) infestations on silage sorghum yield using unmanned aerial systems (UAS)

In the U.S. since 2013, the sugarcane aphid is a perennial pest to all types of sorghum. Rating sugarcane aphid population density, plant damage, and other traits in sorghum requires a large amount of labor and ratings, especially damage ratings, may vary by evaluator. Thus Unmanned Aerial Systems (UAS)–based imagery may be exceedingly useful to more accurately quantify the effects on sorghum caused by sugarcane aphids. This study quantified the dynamic nature of sugarcane aphid infestations on silage sorghum varieties using UAS-based imagery data, and demonstrated the UAS-based measurements correlated to ground measurements. Two UAS platforms equipped with RGB (red, green, and blue) and multispectral cameras respectively were used to evaluate the silage sorghum variety trials during the growing seasons of 2019 and 2020. For the purpose of high throughput phenotyping in sorghum breeding, a new workflow scheme was developed including UAS image processing, raster calculation, DTM (digital terrain model) and CHM (canopy height model) generation, image extraction of sorghum plants, and tabular dataset generation from zonal statistics for further statistical analyses. Ground-based measurements included aphid sampling, aphid damage ratings, plant height, and biomass yields. The normalized difference red edge index (NDRE) and canopy cover collected by the UAS showed negative linear relationship with aphid damage ratings in both trials (R2 = 0.55–0.64). In addition to assessing spatial differences among the varieties in 2019, temporal change in both NDRE and canopy cover from the baseline sampling date in 2020 better estimated aphid damage, R2 of 0.68 and 0.79 respectively, than using the spatial difference of NDRE (R2 = 0.55) and canopy cover (R2 = 0.57) before harvest. Plant height (R2 = 0.84, Root-Mean-Square Error (RMSE) = 0.16 m) can be estimated with efficiency and precision using UAS-derived measurements during high throughput phenotyping of sorghum. Fresh yield estimates for the primary harvests were consistent in both years, but green yield estimates differed among harvests and need to be improved. Future development of UAS-based high throughput phenotyping would benefit from increased temporal resolutions of growth parameters and vegetation indices throughout a growing season.

Digital Terrain Models Generated with Low-Cost UAV Photogrammetry: Methodology and Accuracy

Digital terrain model (DTM) generation is essential to recreating terrain morphology once the external elements are removed. Traditional survey methods are still used to collect accurate geographic data on the land surface. Given the emergence of unmanned aerial vehicles (UAVs) equipped with low-cost digital cameras and better photogrammetric methods for digital mapping, efficient approaches are necessary to allow rapid land surveys with high accuracy. This paper provides a review, complemented with the authors’ experience, regarding the UAV photogrammetric process and field survey parameters for DTM generation using popular commercial photogrammetric software to process images obtained with fixed-wing or multicopter UAVs. We analyzed the quality and accuracy of the DTMs based on four categories: (i) the UAV system (UAV platforms and camera); (ii) flight planning and image acquisition (flight altitude, image overlap, UAV speed, orientation of the flight line, camera configuration, and georeferencing); (iii) photogrammetric DTM generation (software, image alignment, dense point cloud generation, and ground filtering); (iv) geomorphology and land use/cover. For flat terrain, UAV photogrammetry provided a horizontal root mean square error (RMSE) between 1 to 3 × the ground sample distance (GSD) and a vertical RMSE between 1 to 4.5 × GSD, and, for complex topography, a horizontal RMSE between 1 to 7 × GSD and a vertical RMSE between 1.5 to 5 × GSD. Finally, we stress that UAV photogrammetry can provide DTMs with high accuracy when the photogrammetric process variables are optimized.

Self-supervised monocular depth estimation from oblique UAV videos

Unmanned Aerial Vehicles (UAVs) have become an essential photogrammetric measurement as they are affordable, easily accessible and versatile. Aerial images captured from UAVs have applications in small and large scale texture mapping, 3D modelling, object detection tasks, Digital Terrain Model (DTM) and Digital Surface Model (DSM) generation etc. Photogrammetric techniques are routinely used for 3D reconstruction from UAV images where multiple images of the same scene are acquired. Developments in computer vision and deep learning techniques have made Single Image Depth Estimation (SIDE) a field of intense research. Using SIDE techniques on UAV images can overcome the need for multiple images for 3D reconstruction. This paper aims to estimate depth from a single UAV aerial image using deep learning. We follow a self-supervised learning approach, Self-Supervised Monocular Depth Estimation (SMDE), which does not need ground truth depth or any extra information other than images for learning to estimate depth. Monocular video frames are used for training the deep learning model which learns depth and pose information jointly through two different networks, one each for depth and pose. The predicted depth and pose are used to reconstruct one image from the viewpoint of another image utilising the temporal information from videos. We propose a novel architecture with two 2D Convolutional Neural Network (CNN) encoders and a 3D CNN decoder for extracting information from consecutive temporal frames. A contrastive loss term is introduced for improving the quality of image generation. Our experiments are carried out on the public UAVid video dataset. The experimental results demonstrate that our model outperforms the state-of-the-art methods in estimating the depths.

Integrating water-classified returns in DTM generation to increase accuracy of stream delineations and geomorphic analyses

High resolution topographic data has become widely available over the preceding decades and increasingly detailed digital elevation models are aiding in nearly every type of natural-resource related research. Digital terrain models (DTMs), which depict the ground surface topography devoid of vegetation or man-made structures, are particularly helpful in stream-related research. Historically, coarse resolution topographic data (e.g., several meters to tens of meters pixel size) did not afford evaluation of meter scale roughness elements exposed above the water surface within stream channels. The purpose of this study is to demonstrate how the integration of water-classified lidar returns in submeter resolution DTM-development may capture stream corridor topography and be useful for further stream-related research. Four reaches of streams draining the southeastern Blue Ridge Escarpment in southern North Carolina (USA) are assessed for reach positioning, length, and gradient. These parameters are chosen because they are foundational to many other forms of stream analysis (e.g., stream power, normalized channel steepness, chi, and others). Water-assigned lidar returns are included in 0.5- pixel size DTMs and compared to both a 0.5-m DTM generated without use of water returns (i.e., bare-earth) and a pre-processed, hydro-flattened 0.9-m bare-earth DTM. In steep bedrock channels, bare-earth only DTMs result in channels 12–23% shorter than water return integrated DTMs. Observations of stream positioning on DTMs that include water returns and comparisons to orthophotographs suggest a more consistent stream center line in relation to boulders and exposed bedrock within stream channels. Small streams do not benefit from the modified analysis methods because water-classified returns are not present in these channels. Nor do low gradient alluvial channels benefit because these streams tend to lack exposed bedrock or large roughness elements that might divert stream flows. Because so many geomorphic parameters are largely dependent on channel length, these findings have far-reaching implications in ongoing stream-related research. The methods presented here do not require new data collection or technology, but offer simple modifications to processing of existing data and should be considered on other high quality lidar datasets.

The Effect of LiDAR Sampling Density on DTM Accuracy for Areas with Heavy Forest Cover

Laser scanning via LiDAR is a powerful technique for collecting data necessary for Digital Terrain Model (DTM) generation, even in densely forested areas. LiDAR observations located at the ground level can be separated from the initial point cloud and used as input for the generation of a Digital Terrain Model (DTM) via interpolation. This paper proposes a quantitative analysis of the accuracy of DTMs (and derived slope maps) obtained from LiDAR data and is focused on conditions common to most forestry activities (rough, steep terrain with forest cover). Three interpolation algorithms were tested: Inverse Distance Weighted (IDW), Natural Neighbour (NN) and Thin-Plate Spline (TPS). Research was mainly focused on the issue of point data density. To analyze its impact on the quality of ground surface modelling, the density of the filtered data set was artificially lowered (from 0.89 to 0.09 points/m2) by randomly removing point observations in 10% increments. This provides a comprehensive method of evaluating the impact of LiDAR ground point density on DTM accuracy. While the reduction of point density leads to a less accurate DTM in all cases (as expected), the exact pattern varies by algorithm. The accuracy of the LiDAR-derived DTMs is relatively good even when LiDAR sampling density is reduced to 0.40–0.50 points/m2 (50–60 % of the initial point density), as long as a suitable interpolation algorithm is used (as IDW proved to be less resilient to density reductions below approximately 0.60 points/m2). In the case of slope estimation, the pattern is relatively similar, except the difference in accuracy between IDW and the other two algorithms is even more pronounced than in the case of DTM accuracy. Based on this research, we conclude that LiDAR is an adequate method for collecting morphological data necessary for modelling the ground surface, even when the sampling density is significantly reduced.

An improved approach to estimate above-ground volume and biomass of desert shrub communities based on UAV RGB images

Above-ground biomass (AGB) is an essential indicator for assessing ecosystem health and carbon storage in desert shrub-related research. Above-ground volume (AGV) of vegetation is a crucial parameter to estimate the AGB. In unmanned aerial vehicle (UAV) remote sensing, the AGV and AGB are mainly estimated by vegetation feature metrics (for example, spectral indices, textural, and structural metrics). However, there is limited study on the AGV and AGB estimation in desert shrub communities by using UAV, and it is difficult to determine the contribution of these metrics to AGV models under eliminating the influence of background factors. Taking a typical desert shrub area in Inner Mongolia, China as an example, this study develops an improved approach to extracted three types of feature metrics simultaneously using UAV RGB (Red, Green, Blue) images. First, digital orthophoto map (DOM) and digital surface model (DSM) were created through the photogrammetric procedure based on UAV RGB images. Second, the digital terrain model (DTM) for canopy height calculation was generated based on DOM and DSM by object-oriented image binary classification and ground elevation interpolation. Here, we recommended the ENVI Landsat Gap-fill tool to interpolate the ground elevation of vegetation areas. Meanwhile, 21 spectral indices, eight textural metrics, and five structural metrics were extracted. Finally, single-variable and multi-variable commonly used regression models were established based on these metrics and measured AGV with a leave-one-out cross-validation. Results showed that: (1) in the proposed model, the contribution of structural, textural, and spectral metric to shrub AGV models was 86.68, 7.08, and 6.24%, respectively. (2) The horizontal and vertical structural metrics, textural metrics, or spectral indices reflected the one-dimensional change of AGV, which had a saturation effect. (3) The canopy volume, combining the horizontal and vertical characteristics of vegetation canopy, could describe the overall change of AGV and played the most essential role in AGV modelling (R2 = 0.928, relative RMSE = 26.8%). The study findings provide a direct reference in determining suitable vegetation feature metrics for monitoring shrub AGV. The proposed approach for DTM generation and AGV estimation is more efficient, accurate and low-cost than before, and it can be a useful bridge between ground-based investigation and satellite remote sensing.

Generation of digital terrain model from multispectral LiDar using different ground filtering techniques

Multi-spectral LiDAR point clouds provide more information about the terrain surface than mono- wavelength LiDAR, especially in water bodies’ areas. The TITAN LiDAR provided by Teledyne Optech is an example of the multispectral LiDAR system. It has three different wavelengths; mid-infrared (CH-1), near-infrared (CH-2), and green (CH-3) (1.550, 1.064, and 0.532 µm) respectively. This study aims to generate a Digital Terrain Model (DTM) through filtering of Airborne LiDAR point cloud using LiDAR open-source software. Four different ground-filtering algorithms were used; Adaptive TIN (ATIN), Elevation Threshold with Expansion Window (ETEW), Progressive Morphological Algorithms (MORPH), and Maximum Local Slope (SLOPE). These algorithms were executed in Open Source software (Airborne LiDAR Data Processing and Analysis Tools (ALDPAT)). Also, the same data was processed using commercial software (ENVI-LiDAR). The study is focused on the extraction of terrain points from each channel separately and making a visual and quantitative comparison between the performance of the four open-source filtering algorithms and that of the ENVI-LiDAR filter. A quantitative analysis was done by the Cloud to Cloud distance (C2C) to provide the values of standard deviation and RMSE. The results showed that the MORPH and ATIN filtering algorithms gave DTMs similar to that produced by ENVI-LiDAR due to their efficiency in dealing with the forested area. In addition to this MORPH gave better results than ATIN in the removal of non-ground points in coastal barrier islands with a standard deviation of ±93 and ±91 mm for channels −1, 2 respectively. In the third channel, the performance of the SLOPE and the ETEW are well in the water area.

Investigation of the performances of advanced image classification‐based ground filtering approaches for digital terrain model generation

AbstractThe majority of the ground filtering techniques proposed so far use several user‐defined parameter values. Since no standard protocols exist to define these parameters, obtaining the optimum filtering performance is very hard, especially in large‐extent areas with abrupt topography changes. This, of course, reveals the necessity of some more efficient strategies to ease the ground filtering process in such areas. Utilizing classified images for ground filtering purpose may be of help to achieve this. Hence, this study, for the first time in the literature, investigated the performances of the state‐of‐the‐art machine learning algorithms maximum likelihood (ML), artificial neural network (ANN), support vector machines (SVM), and random forest (RF) in ground filtering of a UAS‐based point cloud. The used approaches were based on the assignment of the points corresponding to the ground‐related classes to the ground class. Evaluations showed that the SVM‐based ground filtering approach achieved the optimum filtering result. The SVM‐, ML‐, RF‐, and ANN‐based ground filtering methods achieved the total errors of 13.2%, 16.4%, 19.6%, and 21.9% in the test site, respectively.

МЕТОДИКА АВТОМАТИЗИРОВАННОЙ ФИЛЬТРАЦИИ ДАННЫХ МОБИЛЬНОГО ЛАЗЕРНОГО СКАНИРОВАНИЯ

The result of laser scanning is an array of laser points. The generation of a single point cloud in a given coordinate system is carried out during the registration process at the stage of preliminary field data processing. At this stage it is also often necessary to filter the data. Laser points with an erroneous position are eliminated during the data filleting. The number of erroneous laser points is determined by the of the laser scanner characteristics, surveyed area peculiarities and weather conditions. The devel-opment of methods and algorithms for filtering laser scanning data is carried out based on the analysis of the laser point spatial position and a certain set of additional characteristics, such as intensity value, echo signal, color value. The technique of mobile laser scanning data filtering for the territory of the road passing among the forest and close to individual industrial facilities and building. The main goal of the proposed filtration technique is to obtain data for automatic generation of an accurate digital terrain model. The filtration technique was developed for data acquired under the least favorable con-ditions – in wet weather. Accuracy estimation of generating digital terrain model based on filtered data was carried out.

Experience in digital terrain model generation for hydrodynamic modeling in the Amur River basin

Experience in digital terrain model generation for hydrodynamic modeling in the Amur River basin

High Resolution Digital Terrain Models of Mercury

We refined our Shape from Shading (SfS) algorithm, which has previously been used to create digital terrain models (DTMs) of the Lunar and Martian surfaces, to generate high-resolution DTMs of Mercury from MESSENGER imagery. To adapt the reconstruction procedure to the specific conditions of Mercury and the available imagery, we introduced two methodic innovations. First, we extended the SfS algorithm to enable the 3D-reconstruction from image mosaics. Because most mosaic tiles were acquired at different times and under various illumination conditions, the brightness of adjacent tiles may vary. Brightness variations that are not fully captured by the reflectance model may yield discontinuities at tile borders. We found that the relaxation of the constraint for a continuous albedo map improves the topographic results of an extensive region removing discontinuities at tile borders. The second innovation enables the generation of accurate DTMs from images with substantial albedo variations, such as hollows. We employed an iterative procedure that initializes the SfS algorithm with the albedo map that was obtained by the previous iteration step. This approach converges and yields a reasonable albedo map and topography. With these approaches, we generated DTMs of several science targets such as the Rachmaninoff basin, Praxiteles crater, fault lines, and several hollows. To evaluate the results, we compared our DTMs with stereo DTMs and laser altimeter data. In contrast to coarse laser altimetry tracks and stereo algorithms, which tend to be affected by interpolation artifacts, SfS can generate DTMs almost at image resolution. The root mean squared errors (RMSE) at our target sites are below the size of the horizontal image resolution. For some targets, we could achieve an effective resolution of less than 10 m/pixel, which is the best resolution of Mercury to date. We critically discuss the limitations of the evaluation methodology.