Preprint
Preprints and early-stage research may not have been peer reviewed yet.
To read the file of this research, you can request a copy directly from the authors.

Abstract

Airborne laser scanning and photogrammetry are two main techniques to obtain 3D data representing the object surface. Due to the high cost of laser scanning, we want to explore the potential of using point clouds derived by dense image matching (DIM), as effective alternatives to laser scanning data. We present a framework to evaluate point clouds from dense image matching and derived Digital Surface Models (DSM) based on automatically extracted sample patches. Dense matching error and noise level are evaluated quantitatively at both the local level and whole block level. Experiments show that the optimal vertical accuracy achieved by dense matching is as follows: the mean offset to the reference data is 0.1 Ground Sampling Distance (GSD); the maximum offset goes up to 1.0 GSD. When additional oblique images are used in dense matching, the mean deviation, the variation of mean deviation and the level of random noise all get improved. We also detect a bias between the point cloud and DSM from a single photogrammetric workflow. This framework also allows to reveal inhomogeneity in the distribution of the dense matching errors due to over-fitted BBA network. Meanwhile, suggestions are given on the photogrammetric quality control.

No file available

Request Full-text Paper PDF

To read the file of this research,
you can request a copy directly from the authors.

ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Modern airborne sensors integrate laser scanners and digital cameras for capturing topographic data at high spatial resolution. The capability of penetrating vegetation through small openings in the foliage and the high ranging precision in the cm range have made airborne LiDAR the prime terrain acquisition technique. In the recent years dense image matching evolved rapidly and outperforms laser scanning meanwhile in terms of the achievable spatial resolution of the derived surface models. In our contribution we analyze the inherent properties and review the typical processing chains of both acquisition techniques. In addition, we present potential synergies of jointly processing image and laser data with emphasis on sensor orientation and point cloud fusion for digital surface model derivation. Test data were concurrently acquired with the RIEGL LMS-Q1560 sensor over the city of Melk, Austria, in January 2016 and served as basis for testing innovative processing strategies. We demonstrate that (i) systematic effects in the resulting scanned and matched 3D point clouds can be minimized based on a hybrid orientation procedure, (ii) systematic differences of the individual point clouds are observable at penetrable, vegetated surfaces due to the different measurement principles, and (iii) improved digital surface models can be derived combining the higher density of the matching point cloud and the higher reliability of LiDAR point clouds, especially in the narrow alleys and courtyards of the study site, a medieval city.
Article
Full-text available
The paper reports some comparisons between commercial software able to automatically process image datasets for 3D reconstruction purposes. The main aspects investigated in the work are the capability to correctly orient large sets of image of complex environments, the metric quality of the results, replicability and redundancy. Different datasets are employed, each one featuring a diverse number of images, GSDs at cm and mm resolutions, and ground truth information to perform statistical analyses of the 3D results. A summary of (photogrammetric) terms is also provided, in order to provide rigorous terms of reference for comparisons and critical analyses.
Article
Full-text available
Accurate 3D city models represent an important source of geospatial information to support various “smart city” applications, such as space management, energy assessment, 3D cartography, noise and pollution mapping as well as disaster management. Even though remarkable progress has been made in recent years, there are still many open issues, especially when it comes to the 3D modelling of complex urban scenarios like historical and densely-built city centres featuring narrow streets and non-conventional building shapes. Most approaches introduce strong building priors/constraints on symmetry and roof typology that penalize urban environments having high variations of roof shapes. Furthermore, although oblique photogrammetry is rapidly maturing, the use of slanted views for façade reconstruction is not completely included in the reconstruction pipeline of state-of-the-art software. This paper aims to investigate state-of-the-art methods for 3D building modelling in complex urban scenarios with the support of oblique airborne images. A reconstruction approach based on roof primitives fitting is tested. Oblique imagery is then exploited to support the manual editing of the generated building models. At the same time, mobile mapping data are collected at cm resolution and then integrated with the aerial ones. All approaches are tested on the historical city centre of Bergamo (Italy).
Article
Full-text available
Digital surface models (DSMs) derived from spaceborne and airborne sensors enable the monitoring of the vertical structures for forests in large areas. Nevertheless, due to the lack of an objective performance assessment for this task, it is difficult to select the most appropriate data source for DSM generation. In order to fill this gap, this paper performs change detection analysis including forest decrease and tree growth. The accuracy of the DSMs is evaluated by comparison with measured tree heights from inventory plots (field data). In addition, the DSMs are compared with LiDAR data to perform a pixel-wise quality assessment. DSMs from four different satellite stereo sensors (ALOS/PRISM, Cartosat-1, RapidEye and WorldView-2), one satellite InSAR sensor (TanDEM-X), two aerial stereo camera systems (HRSC and UltraCam) and two airborne laser scanning datasets with different point densities are adopted for the comparison. The case study is a complex central European temperate forest close to Traunstein in Bavaria, Germany. As a major experimental result, the quality of the DSM is found to be robust to variations in image resolution, especially when the forest density is high. The forest decrease results confirm that besides aerial photogrammetry data, very high resolution satellite data, such as WorldView-2, can deliver results with comparable quality as the ones derived from LiDAR, followed by TanDEM-X and Cartosat DSMs. The quality of the DSMs derived from ALOS and Rapid-Eye data is lower, but the main changes are still correctly highlighted. Moreover, the vertical tree growth and their relationship with tree height are analyzed. The major tree height in the study site is between 15 and 30 m and the periodic annual increments (PAIs) are in the range of 0.30–0.50 m.
Article
Full-text available
During the last decade the use of airborne multi camera systems increased significantly. The development in digital camera technology allows mounting several mid- or small-format cameras efficiently onto one platform and thus enables image capture under different angles. Those oblique images turn out to be interesting for a number of applications since lateral parts of elevated objects, like buildings or trees, are visible. However, occlusion or illumination differences might challenge image processing. From an image orientation point of view those multi-camera systems bring the advantage of a better ray intersection geometry compared to nadir-only image blocks. On the other hand, varying scale, occlusion and atmospheric influences which are difficult to model impose problems to the image matching and bundle adjustment tasks. In order to understand current limitations of image orientation approaches and the influence of different parameters such as image overlap or GCP distribution, a commonly available dataset was released. The originally captured data comprises of a state-of-the-art image block with very high overlap, but in the first stage of the so-called ISPRS/EUROSDR benchmark on multi-platform photogrammetry only a reduced set of images was released. In this paper some first results obtained with this dataset are presented. They refer to different aspects like tie point matching across the viewing directions, influence of the oblique images onto the bundle adjustment, the role of image overlap and GCP distribution. As far as the tie point matching is concerned we observed that matching of overlapping images pointing to the same cardinal direction, or between nadir and oblique views in general is quite successful. Due to the quite different perspective between images of different viewing directions the standard tie point matching, for instance based on interest points does not work well. How to address occlusion and ambiguities due to different views onto objects is clearly a non-solved research problem so far. In our experiments we also confirm that the obtainable height accuracy is better when all images are used in bundle block adjustment. This was also shown in other research before and is confirmed here. Not surprisingly, the large overlap of 80/80% provides much better object space accuracy – random errors seem to be about 2-3fold smaller compared to the 60/60% overlap. A comparison of different software approaches shows that newly emerged commercial packages, initially intended to work with small frame image blocks, do perform very well.
Article
Full-text available
Abstract In the past two decades Object-Based Image Analysis (OBIA) established itself as an efficient approach for the classification and extraction of information from remote sensing imagery and, increasingly, from non-image based sources such as Airborne Laser Scanner (ALS) point clouds. ALS data is represented in the form of a point cloud with recorded multiple returns and intensities. In our work, we combined OBIA with ALS point cloud data in order to identify and extract buildings as 2D polygons representing roof outlines in a top down mapping approach. We performed rasterization of the ALS data into a height raster for the purpose of the generation of a Digital Surface Model (DSM) and a derived Digital Elevation Model (DEM). Further objects were generated in conjunction with point statistics from the linked point cloud. With the use of class modelling methods, we generated the final target class of objects representing buildings. The approach was developed for a test area in Biberach an der Riß (Germany). In order to point out the possibilities of the adaptation-free transferability to another data set, the algorithm has been applied “as is” to the ISPRS Benchmarking data set of Toronto (Canada). The obtained results show high accuracies for the initial study area (thematic accuracies of around 98%, geometric accuracy of above 80%). The very high performance within the ISPRS Benchmark without any modification of the algorithm and without any adaptation of parameters is particularly noteworthy.
Article
Full-text available
For monitoring purposes and in the context of geomorphological research, Unmanned Aerial Vehicles (UAV) appear to be a promising solution to provide multi-temporal Digital Surface Models (DSMs) and orthophotographs. There are a variety of photogrammetric software tools available for UAV-based data. The objective of this study is to investigate the level of accuracy that can be achieved using two of these software tools: Agisoft PhotoScan® Pro and an open-source alternative, IGN© MicMac®, in sub-optimal survey conditions (rugged terrain, with a large variety of morphological features covering a range of roughness sizes, poor GPS reception). A set of UAV images has been taken by a hexacopter drone above the Rivière des Remparts, a river on Reunion Island. This site was chosen for its challenging survey conditions: the topography of the study area (i) involved constraints on the flight plan; (ii) implied errors on some GPS measurements; (iii) prevented an optimal distribution of the Ground Control Points (GCPs) and; (iv) was very complex to reconstruct. Several image processing tests are performed with different scenarios in order to analyze the sensitivity of each software package to different parameters (image quality, numbers of GCPs, etc.). When computing the horizontal and vertical errors within a control region on a set of ground reference targets, both methods provide rather similar results. A precision up to 3–4 cm is achievable with these software packages. The DSM quality is also assessed over the entire study area comparing PhotoScan DSM and MicMac DSM with a Terrestrial Laser Scanner (TLS) point cloud. PhotoScan and MicMac DSM are also compared at the scale of particular features. Both software packages provide satisfying results: PhotoScan is more straightforward to use but its source code is not open; MicMac is recommended for experimented users as it is more flexible.
Article
Full-text available
Novel automated photogrammetry is based on four innovations. First is the cost-free increase of overlap between images when sensing digitally. Second is an improved radiometry. Third is multi-view matching. Fourth is the Graphics Processing Unit (GPU), making complex algorithms for image matching very practical. These innovations lead to improved automation of the photogrammetric workflow so that point clouds are created at sub-pixel accuracy, at very dense intervals, and in near real-time, thereby eroding the unique selling proposition of lidar scanners. Two test projects compare point clouds from aerial and street-side lidar systems with those created from images. We show that the photogrammetric accuracy compares well with the lidar-method, yet the density of surface points is much higher from images. and the throughput is commensurate with a fully automated all-digital approach. Beyond this density, we identify 15 additional advantages of the photogrammetric approach.
Conference Paper
Full-text available
This article presents an approach of a multi-view stereo (MVS) method for the generation of dense and precise 3D point clouds. It is based on the Semi-Global Matching (SGM) method followed by a fusion step in which the redundant depth estimations across single stereo models are merged. We present a hierarchical coarse-to-fine solution for the SGM method in which matching results of low resolution pyramids are used to limit disparity search ranges for high resolution pyramids. By means of large format aerial and close range imagery we show that memory demands as well as processing times can be significantly decreased whereas the quality of resulting disparities estimations is maintained. By merging redundant disparity estimations of multiple stereo models the precision and robustness of the generated point clouds can be increased. Based on basic principles of epipolar geometry we present a time efficient algorithm for outlier detection and object point triangulation minimizing the reprojection error. Thereby the geometric configuration of adjacent cameras is taken into account. An implementation of the algorithm called SURE as well the library interface libTSgm of the presented algorithm is publicly available at: http://www.ifp.uni-stuttgart.de/publications/software/.
Article
Full-text available
High density point clouds of urban scenes are used to identify object classes like buildings, vegetation, vehicles, ground, and water. Point cloud segmentation can support classification and further feature extraction provided that the segments are logical groups of points belonging to the same object class. A single segmentation method will typically not provide a satisfactory segmentation for a variety of classes. This paper explores the combination of various segmentation and post-processing methods to arrive at useful point cloud segmentations. A feature based on the normal vector and flatness of a point neighbourhood is used to group cluttered points in trees as well as points on surfaces in areas where the extraction of planes was not successful. Combined with segment merging and majority filtering large segments can be obtained allowing the derivation of accurate segment feature values. Results are presented and discussed for a 70 million point dataset over a part of Rotterdam.
Article
Full-text available
The complementation of terrestrial laser scanning (TLS) by photogrammetric acquisition techniques enables the exploitation of two different measurement principles. TLS with distance measurements can be used for acquiring large-scale point clouds at medium range distances, while image-based surface reconstruction methods enable flexible acquisition with high precision at short distances. Within this paper, an automatic procedure for the combination of photogrammetry and laser scanning is presented. Synthetic images derived from the laser scanning point clouds are used together with camera images in a common Structure-from-Motion (SfM) process. This enables an automatic registration of multiple laser scans. Furthermore, surface information can also be derived from the imagery using dense image matching methods. While the laser scanning data provides the missing scale information, the imagery can complement the dataset to fill gaps, occlusions or to resolve small details. Consequently, large datasets can be connected by registering imagery and laser scans.
Article
Full-text available
Until recently, the acquisition of high quality digital elevation models was dominated by the use of airborne lidar. However, meanwhile the increasing quality of digital airborne cameras in combination with recent improvements in matching algorithms allow for an automatic image-based acquisition as a suitable alternative. Within the paper, this progress will be demonstrated on the example of photogrammetric DEM generation using the semi-global matching (SGM) stereo method. Since this approach aims at pixel-wise matching, dense 3D point clouds can be generated. The tests described in the paper are based on data collected from different digital airborne cameras at various flight scenarios during a recent test on photogrammetric 3D data capture. Thus, the impact of different stereo configurations on the quality of the final outcome can be evaluated and compared to test results already available. Special interest is also paid to the combination of multiple stereo pairs with different base-to-height ratios, which can be used efficiently to increase the precision and reliability of the matching results. German Die Erzeugung digitaler Höhenmodelle wurde bis vor kurzem durch den Einsatz von Lidar dominiert. Durch die verbesserte Qualität von Kamerasystemen und die voranschreitende Entwicklung bildbasierter Zuordnungsverfahren stellt die Datenerfassung mittels Luftbildkameras mittlerweile eine ernstzunehmende Alternative dar. In diesem Artikel wird dieser Fortschritt anhand des Beispiels der photogrammetrischen DEM Generierung durch Semi-Global Matching (SGM) aufgezeigt. Mittels dieser Methode werden pixelweise Zuordnungen zwischen zwei Bildpaaren erhalten, so dass extrem dichte Oberflächenmodelle generiert werden können. Für die in diesem Artikel durchgeführten Tests wurde Bildmaterial unterschiedlicher Kamerasysteme und Flugkon gurationen untersucht. Dabei wurde unter anderem auf Daten im Rahmen kürzlich durchgeführter Tests zur Untersuchung photogrammetrischer Datenerfassung zurückgegriffen. So kann der Ein uss von ugspezi schen Parametern auf die erhaltenen Resultate evaluiert, und selbige mit bestehenden Ergebnissen verglichen werden. Spezielle Beachtung ndet außerdem die Fusion mehrerer Stereomodelle resultierend aus Bildpaaren unterschiedlicher Basis-Höhenverhältnisse. Dies ermöglicht eine Steigerung hinsichtlich der Genauigkeit und der geometrischen Qualität.
Article
Full-text available
There have been increasing demands for automatically monitoring urban areas in very high detail, and the Unmanned Aerial Vehicle (UAV) with auto-navigation (AUNA) system offers such capability. This study proposes an object-based hierarchical method to detect changes from UAV images taken at different times. It consists of several steps. In the first step, an octocopter with AUNA capability is used to acquire images at different dates. These images are registered automatically, based on SIFT (Scale-Invariant Feature Transform) feature points, via the general bundle adjustment framework. Thus, the Digital Surface Models (DSMs) and orthophotos can be generated for raster-based change analysis. In the next step, a multi-primitive segmentation method combining the spectral and geometric information is proposed for object-based analysis. In the final step, a multi-criteria decision analysis is carried out concerning the height, spectral and geometric coherence, and shape regularity for change determination. Experiments based on UAV images with five-centimeter ground resolution demonstrate the effectiveness of the proposed method, leading to the conclusion that this method is practically applicable for frequent monitoring.
Article
Full-text available
During the implementation of the DGPF-project on Digital Photogrammetric Camera Evaluation a team “Digital Elevation Models” was established. The main goal was to use the test's framework for documentation and evaluation of the current state-of-the-art on photogrammetric 3D data capture from automatic image matching. During these investigations the accuracy and reliability of DSM rasters and 3D point clouds as derived from imagery of digital photogrammetric camera systems were evaluated. For this purpose they were compared to reference measurements from ground truth and airborne LiDAR. In addition to the evaluation of standard products, the usability of elevation data from image matching was investigated while aiming at specific applications in the context of urban modeling and forestry. German Während des DGPF-Projektes zur Evaluierung digitaler photogrammetrischer Luftbildkamerasysteme wurde auch eine Auswertegruppe für die Bewertung der Genauigkeit der Höhenmodellgenerierung etabliert. Dabei sollte der DGPF-Test genutzt werden, um den derzeitigen Stand der Technik der photogrammetrischen 3D Erfassug mittels automatischer Bildzuordnung zu dokumentieren. Hierfür wurden DSM Raster und 3D Punktwolken aus Bildern der photgrammetrischen Kamerasysteme abgeleitetet und die Qualität dieser Ergebnisse in Bezug auf Genauigkeit und Zuverlässigkeit bewertet. Dabei wurde ein Vergleich zu terrestrischen Referenzmessungen und flugzeuggestützen LiDAR Daten durchgeführt. Neben der qualitativen Bewertung von Standardprodukten wurde auch die Nutzbarkeit der Höhendaten für spezielle Anwendungen beispielsweise im Kontext der 3D Stadmodellierung und Forstwirtschaft untersucht.
Article
Full-text available
Digital surface models (DSMs) are widely used in forest science to model the forest canopy. Stereo pairs of very high resolution satellite and digital aerial images are relatively new and their absolute accuracy for DSM generation is largely unknown. For an assessment of these input data two DSMs based on a WorldView-2 stereo pair and a ADS80 DSM were generated with photogrammetric instruments. Rational polynomial coefficients (RPCs) are defining the orientation of the WorldView-2 satellite images, which can be enhanced with ground control points (GCPs). Thus two WorldView-2 DSMs were distinguished: a WorldView-2 RPCs-only DSM and a WorldView-2 GCP-enhanced RPCs DSM. The accuracy of the three DSMs was estimated with GPS measurements, manual stereo-measurements, and airborne laser scanning data (ALS). With GCP-enhanced RPCs the WorldView-2 image orientation could be optimised to a root mean square error (RMSE) of 0.56 m in planimetry and 0.32 m in height. This improvement in orientation allowed for a vertical median error of -0.24 m for the WorldView-2 GCP-enhanced RPCs DSM in flat terrain. Overall, the DSM based on ADS80 images showed the highest accuracy of the three models with a median error of 0.08 m over bare ground. As the accuracy of a DSM varies with land cover three classes were distinguished: herb and grass, forests, and artificial areas. The study suggested the ADS80 DSM to best model actual surface height in all three land cover classes, with median errors <1.1 m. The WorldView-2 GCP-enhanced RPCs model achieved good accuracy, too, with median errors of -0.43 m for the herb and grass vegetation and -0.26 m for artificial areas. Forested areas emerged as the most difficult land cover type for height modelling; still, with median errors of -1.85 m for the WorldView-2 GCP-enhanced RPCs model and -1.12 m for the ADS80 model, the input data sets evaluated here are quite promising for forest canopy modelling.
Article
Full-text available
With the increasing pulse rates of airborne laser scanners point clouds become an interesting data source for mapping beyond the production of digital terrain models. Whereas height accuracy of point clouds is of prime importance for terrain models, mapping of buildings and other objects also requires an analysis of planimetric accuracy of point clouds. This paper presents a methodology to estimate relative strip offsets and standard deviations of X-and Y-coordinates using automatically extracted ridge lines of buildings in overlaps between strips. An analysis of three blocks with 200-250 million points each shows that constant strip offsets explain the majority of planimetric shifts between ridge lines in strip overlaps. It is demonstrated that strip offsets below 3-4 cm and platform positioning noise of only 2 cm can be obtained with GPS configurations consisting of two onboard receivers and multiple reference stations in the terrain and a low flying height.
Article
LiDAR has been an effective technology for acquiring urban land cover data in recent decades. Previous studies indicate that geometric features have a strong impact on land cover classification. Here, we analyzed an urban LiDAR dataset to explore the optimal feature subset from 25 geometric features incorporating 25 scales under 6 definitions for urban land cover classification. We performed a feature selection strategy to remove irrelevant or redundant features based on the correlation coefficient between features and classification accuracy of each features. The neighborhood scales were divided into small (0.5–1.5 m), medium (1.5–6 m) and large (>6 m) scale. Combining features with lower correlation coefficient and better classification performance would improve classification accuracy. The feature depicting homogeneity or heterogeneity of points would be calculated at a small scale, and the features to smooth points at a medium scale and the features of height different at large scale. As to the neighborhood definition, cuboid and cylinder were recommended. This study can guide the selection of optimal geometric features with adaptive neighborhood scale for urban land cover classification.
Conference Paper
This study aims to highlight the differences, in terms of robustness and efficiency, of the use of LIDAR point clouds compared to dense image matching (DIM) point clouds at urban areas that contain buildings with complex structure. The application is conducted over an area in the Greek island of Milos using two different types of data: (a) a dense point cloud which extracted by DIM using a variation of the stereo-method semi-global matching (SGM) at RGB digital aerial images, and (b) a georeferenced LIDAR point cloud. For the case of the DIM point cloud, the following steps were applied: aerial triangulation, rectification of the original images to epipolar images, extraction of disparity maps and application of a 3D similarity transformation. The evaluations that were executed included urban and rural areas. At first step, a direct cloud-to-cloud comparison between the georeferenced DIM and LIDAR point clouds was carried out. Then, the corresponding orthoimages generated by the DIM and LIDAR point clouds undergo a quality control. Although the results show that the LIDAR point clouds respond better at such complex scenes compared to DIM point clouds, the latter gave promising results. In this context, the Quality Assurance issue is also discussed so as to be more efficient towards the challenge of the increasingly greater demands for accurate and cost effective applications.
Article
In this article the performance of dense image matching (DIM) is investigated regarding its capability to yield terrain data, especially close to free-flowing water ways. Therefore, over two study areas aerial images with ground sampling distances 10 cm and 6 cm, respectively, are used for matching with the software packages Match-T (Trimble) and SURE (nframes). The matching results over areas with varying vegetation density (open grassland, loose and dense vegetation) are then compared with ALS reference data. Two parameters are investigated: (a) the terrain coverage; i.e. the percentage of the terrain covered by the matching results; and (b) the height accuracy of the matching results in the terrain class. The results show that DIM can only deliver terrain data in areas with no or very loose vegetation. Additionally, it was found that in the case of open grassland the DIM terrain heights were systematically higher by 10cm compared with the ALS terrain heights. This is caused by the fact that ALS can penetrate the vegetation to some extent whereas matching occurs on top of the grass. The very good height accuracy (as standard deviation) obtainable by DIM, which is only slightly worse than the ALS accuracy (6.5 cm vs. 4.5 cm), is encouraging. Motivated by these results new possible applications arise for the respective authorities (the German Federal Institute of Hydrology and the German Federal Wa ter and Shipping Administration), e.g. capturing dry fallen areas of free flowing rivers during documentation at low water levels. © 2016 E. Schweizerbart'sche Verlagsbuchhandlung, Stuttgart, Germany.
Article
3D scene analysis in terms of automatically assigning 3D points a respective semantic label has become a topic of great importance in photogrammetry, remote sensing, computer vision and robotics. In this paper, we address the issue of how to increase the distinctiveness of geometric features and select the most relevant ones among these for 3D scene analysis. We present a new, fully automated and versatile framework composed of four components: (i) neighborhood selection, (ii) feature extraction, (iii) feature selection and (iv) classification. For each component, we consider a variety of approaches which allow applicability in terms of simplicity, efficiency and reproducibility, so that end-users can easily apply the different components and do not require expert knowledge in the respective domains. In a detailed evaluation involving 7 neighborhood definitions, 21 geometric features, 7 approaches for feature selection, 10 classifiers and 2 benchmark datasets, we demonstrate that the selection of optimal neighborhoods for individual 3D points significantly improves the results of 3D scene analysis. Additionally, we show that the selection of adequate feature subsets may even further increase the quality of the derived results while significantly reducing both processing time and memory consumption.
Article
Use of direct geo-referencing data leads to registration failure between sequent images and LiDAR data captured by mini-UAV platforms because of low-cost sensors. This paper therefore proposes a novel automatic registration method for sequent images and LiDAR data captured by mini-UAVs. First, the proposed method extracts building outlines from LiDAR data and images and estimates the exterior orientation parameters (EoPs) of the images with building objects in the LiDAR data coordinate framework based on corresponding corner points derived indirectly by using linear features. Second, the EoPs of the sequent images in the image coordinate framework are recovered using a structure from motion (SfM) technique, and the transformation matrices between the LiDAR coordinate and image coordinate frameworks are calculated using corresponding EoPs, resulting in a coarse registration between the images and the LiDAR data. Finally, 3D points are generated from sequent images by multi-view stereo (MVS) algorithms. Then the EoPs of the sequent images are further refined by registering the LiDAR data and the 3D points using an iterative closest-point (ICP) algorithm with the initial results from coarse registration, resulting in a fine registration between sequent images and LiDAR data. Experiments were performed to check the validity and effectiveness of the proposed method. The results show that the proposed method achieves high-precision robust co-registration of sequent images and LiDAR data captured by mini-UAVs.
Article
3D building models, being the main part of a digital city scene, are essential to all applications related to human activities in urban environments. The development of range sensors and Multi-View Stereo (MVS) technology facilitates our ability to automatically reconstruct level of details 2 (LoD2) models of buildings. However, because of the high complexity of building structures, no fully automatic system is currently available for producing building models. In order to simplify the problem, a lot of research focuses only on particular buildings shapes, and relatively simple ones. In this paper, we analyze the property of topology graphs of object surfaces, and find that roof topology graphs have three basic elements: loose nodes, loose edges, and minimum cycles. These elements have interesting physical meanings: a loose node is a building with one roof face; a loose edge is a ridge line between two roof faces whose end points are not defined by a third roof face; and a minimum cycle represents a roof corner of a building. Building primitives, which introduce building shape knowledge, are defined according to these three basic elements. Then all buildings can be represented by combining such building primitives. The building parts are searched according to the predefined building primitives, reconstructed independently, and grouped into a complete building model in a CSG-style. The shape knowledge is inferred via the building primitives and used as constraints to improve the building models, in which all roof parameters are simultaneously adjusted. Experiments show the flexibility of building primitives in both lidar point cloud and stereo point cloud.
Article
Image matching has a history of more than 50 years, with the first experiments performed with analogue procedures for cartographic and mapping purposes. The recent integration of computer vision algorithms and photogrammetric methods is leading to interesting procedures which have increasingly automated the entire image-based 3D modelling process. Image matching is one of the key steps in 3D modelling and mapping. This paper presents a critical review and analysis of four dense image-matching algorithms, available as open-source and commercial software, for the generation of dense point clouds. The eight datasets employed include scenes recorded from terrestrial and aerial blocks, acquired with convergent and normal (parallel axes) images, and with different scales. Geometric analyses are reported in which the point clouds produced with each of the different algorithms are compared with one another and also to ground-truth data.
Article
Recent research results have shown that the performance of digital surface model extraction using novel high-quality photogrammetric images and image matching is a highly competitive alternative to laser scanning. In this article, we proceed to compare the performance of these two methods in the estimation of plot-level forest variables. Dense point clouds extracted from aerial frame images were used to estimate the plot-level forest variables needed in a forest inventory covering 89 plots. We analyzed images with 60% and 80% forward overlaps and used test plots with off-nadir angles of between 0° and 20°. When compared to reference ground measurements, the airborne laser scanning (ALS) data proved to be the most accurate: it yielded root mean square error (RMSE) values of 6.55% for mean height, 11.42% for mean diameter, and 20.72% for volume. When we applied a forward overlap of 80%, the corresponding results from aerial images were 6.77% for mean height, 12.00% for mean diameter, and 22.62% for volume. A forward overlap of 60% resulted in slightly deteriorated RMSE values of 7.55% for mean height, 12.20% for mean diameter, and 22.77% for volume. According to our results, the use of higher forward overlap produced only slightly better results in the estimation of these forest variables. Additionally, we found that the estimation accuracy was not significantly impacted by the increase in the off-nadir angle. Our results confirmed that digital aerial photographs were about as accurate as ALS in forest resources estimation as long as a terrain model was available.
Article
Airborne laser scanning and image matching can today form the basis for the generation of digital terrain models (DTMs). In addition to the DTM, quality parameters are needed that describe the accuracy at a high level of detail, at best for every interpolated DTM point. Furthermore, other parameters are of interest, for example, the distance from each DTM point to the data point next to it. This paper presents a method to derive accuracy measures from the original data and the DTM itself. Its application is demonstrated with an example. The quality measures are suitable for informing users in detail about DTM quality and warning them of weakly determined areas. © 2006 The Authors. Journal Compilation 2006 The Remote Sensing and Photogrammetry Society and Blackwell Publishing Ltd.
Article
The growing applications of digital surface models (DSMs) for object detection, segmentation and representation of terrestrial landscapes have provided impetus for further automation of 3D spatial information extraction processes. While new technologies such as lidar are available for almost instant DSM generation, the use of stereoscopic high-resolution satellite imagery (HRSI), coupled with image matching, affords cost-effective measurement of surface topography over large coverage areas. This investigation explores the potential of IKONOS Geo stereo imagery for producing DSMs using an alternative sensor orientation model, namely bias-corrected rational polynomial coefficients (RPCs), and a hybrid image-matching algorithm. To serve both as a reference surface and a basis for comparison, a lidar DSM was employed in the Hobart testfield, a region of differing terrain types and slope. In order to take topographic variation within the modelled surface into account, the lidar strip was divided into separate sub-areas representing differing land cover types. It is shown that over topographically diverse areas, heighting accuracy to better than 3 pixels can be readily achieved. Results improve markedly in feature-rich open and relatively flat terrain, with sub-pixel accuracy being achieved at check points surveyed using the global positioning system (GPS). This assessment demonstrates that the outlook for DSM generation from HRSI is very promising. © 2005 The Remote Sensing and Photogrammetry Society and Blackwell Publishing Ltd.
Conference Paper
Automatic detection of damaged buildings from aerial and satellite images is an important problem for rescue planners and military personnel. In this study, we present a novel approach for automatic detection of damaged buildings in color aerial images. Our method is based on color invariants for building rooftop segmentation. Then, we benefit from grayscale histogram to extract shadow segments. After building verification using shadow information, we define a new damage measure for each building. Experimentally, we show that using our damage measure it is possible to discriminate nearby damaged and undamaged buildings. We present our experimental results on aerial images.
Article
A comparison between data acquisition and processing from passive optical sensors and airborne laser scanning is presented. A short overview and the major differences between the two technologies are outlined. Advantages and disadvantages with respect to various aspects are discussed, like sensors, platforms, flight planning, data acquisition conditions, imaging, object reflectance, automation, accuracy, flexibility and maturity, production time and costs. A more detailed comparison is presented with respect to DTM and DSM generation. Strengths of laser scanning with respect to certain applications are outlined. Although airborne laser scanning competes to a certain extent with photogrammetry and will replace it in certain cases, the two technologies are fairly complementary and their integration can lead to more accurate and complete products, and open up new areas of application.
Article
Measures for the accuracy assessment of Digital Elevation Models (DEMs) are discussed and characteristics of DEMs derived from laser scanning and automated photogrammetry are presented. Such DEMs are very dense and relatively accurate in open terrain. Built-up and wooded areas, however, need automated filtering and classification in order to generate terrain (bare earth) data when Digital Terrain Models (DTMs) have to be produced. Automated processing of the raw data is not always successful. Systematic errors and many outliers at both methods (laser scanning and digital photogrammetry) may therefore be present in the data sets. We discuss requirements for the reference data with respect to accuracy and propose robust statistical methods as accuracy measures. Their use is illustrated by application at four practical examples. It is concluded that measures such as median, normalized median absolute deviation, and sample quantiles should be used in the accuracy assessment of such DEMs. Furthermore, the question is discussed how large a sample size is needed in order to obtain sufficiently precise estimates of the new accuracy measures and relevant formulae are presented.
Benchmarking high density image matching for oblique airborne imagery
  • S Cavegn
  • N Haala
  • S Nebiker
  • M Rothermel
  • P Tutzauer
Cavegn, S., Haala, N., Nebiker, S., Rothermel, M. and Tutzauer, P., 2014. Benchmarking high density image matching for oblique airborne imagery. Int. Arch. Photogram. Remote Sens. Spatial Inf. Sci. 40(3), 45-52.
Semi-global matching: An alternative to LIDAR for DSM generation
  • S Gehrke
  • K Morin
  • M Downey
  • N Boehrer
  • T Fuchs
Gehrke, S., Morin, K., Downey, M., Boehrer, N. and Fuchs, T., 2010. Semi-global matching: An alternative to LIDAR for DSM generation. Proceedings of the 2010 Canadian Geomatics Conf. and Symp. of Commission I. 1-6.
Benchmark on image matching final report
  • N Haala
Haala, N., 2015. Benchmark on image matching final report. EuroSDR Official Publication. 115-145.
ISPRS benchmark for multiplatform photogrammetry
  • F Nex
  • M Gerke
  • F Remondino
  • H J Przybilla
  • M Bäumker
  • A Zurhorst
Nex, F., Gerke, M., Remondino, F., Przybilla, H.J., Bäumker, M. and Zurhorst, A., 2015. ISPRS benchmark for multiplatform photogrammetry. ISPRS Ann. Photogram. Remote Sens. Spatial Inf. Sci. 2(3), 135-142.
Potential of dense matching for the generation of high quality digital elevation models. ISPRS Hannover Workshop for High-Resolution Earth Imaging for Geospatial Information
  • M Rothermel
  • N Haala
Rothermel, M. and Haala, N., 2011. Potential of dense matching for the generation of high quality digital elevation models. ISPRS Hannover Workshop for High-Resolution Earth Imaging for Geospatial Information. 331-343.
Dense matching quality evaluation-an empirical study
  • Z Zhang
  • M Gerke
  • M Peter
  • M Y Yang
  • G Vosselman
Zhang, Z., Gerke, M., Peter, M., Yang, M.Y. and Vosselman, G., 2017. Dense matching quality evaluation-an empirical study. IEEE Joint Urban Remote Sensing Event (JURSE). 1-4.