Research
Homepage

Photogrammetric Computer Vision

Research Activities and Projects


Dense 3D Surface Reconstruction by Aerial Image Matching Dense 3D Surface Reconstruction by Aerial Image Matching

For a considerable period, the acquisition of high quality Digital Elevation Models (DEM) was dominated by the use of airborne LiDAR. Meanwhile, automatic image based collection revived as a suitable alternative. This development was triggered by the increasing quality of digital airborne cameras as well as recent innovations in matching algorithms. Airborne imagery of good dynamic and signal-to-noise-ratio as available from digital aerial cameras is highly beneficial for automatic image matching. This is especially true for surfaces with relatively little surface texture. Consequently, the quality and accuracy of image based point transfer as basic observation for 3D surface reconstruction ameliorated considerably. Recent tests have already demonstrated the feasibility of image matching as a valid alternative to airborne LiDAR. More ...


Unmanned Airborne Systems/Vehicles (UAS/UAV) Unmanned Airborne Systems/Vehicles (UAS/UAV)

UAVs are unmanned aerial vehicles which can fly autonomously, semi-autonomously or manually steered by a pilot from the ground using a remote control. Since these systems can be completely built-up at very reasonable prices, their use as mapping platform is very cost effective. For mapping UAVs are equipped with photogrammetric measurement systems like small or medium size cameras . Depending on the payload of the UAV, thermal or infrared camera systems as well as multispectral cameras can be integrated additionally. For the determination of the trajectory, UAVs additionally feature an integrated GNSS/INS system (global navigation satellite system / inertial navigation system), barometric altimeter and compass. More ...


3D UAS Flight Planning for Building Reconstruction 3D UAS Flight Planning for Building Reconstruction

In in many cases, the task of acquiring imagery for building / fašade reconstruction can only be solved unsatisfyingly from terrestrial positions. The camera station configuration is usually disadvantageous for higher building areas. Additionally, many areas may remain unobserved due to dead angles. To overcome these deficiencies, the use of an UAS for image acquisition can be helpful. However, tools to automate a corresponding flight exist only for structures of low complexity. As a result, expert pilots are needed to maneuver and collect suited data. More ...


Benchmark on High Density Aerial Image Matching Image-Based Mobile Mapping

Dense sequences of street level imagery as captured from multi-view stereo mobile mapping systems provide a valuable source of information e.g. during monitoring of urban infrastructures. Dense multi-view stereo image matching can successfully be applied for efficient 3D data collection even in complex and densely built-up environments. However, the extraction of 3D point clouds or depth maps from such street-level imagery presumes high quality georeferencing as a first processing step. In densely built-up areas and urban canyons, poor GNSS coverage frequently prevents sufficient accuracy and reliability. This joint project aims on the development of an image-based georeferencing approach, which additionally integrates tie and control point information and thus enables a geometric evaluation for this street level mobile mapping image sequences at a high quality. More ...


Virtual City Models Virtual City Models

The development of tools for the generation of 3D city models started almost two decades ago. From the beginning, fully automatic reconstruction systems were envisioned to fulfill the need for efficient data collection. However, research on automatic city modelling is still an active area. Automatic city modelling aims at the 3D reconstruction of polyhedral building objects, which mainly reflects the respective roof shapes and building footprints. For this purpose, airborne images or laser scans are used. In addition, more detailed facade geometries are generated from terrestrial data collection. More ...


Benchmark on High Density Aerial Image Matching Benchmark on High Density Aerial Image Matching

Improvements in camera technology and the availability pixel-wise matching approaches triggered the development of new software tools for image based 3D reconstruction. Meanwhile research groups as well as commercial vendors provide photogrammetric software to generate dense, reliable and accurate 3D point clouds and Digital Surface Models (DSM) from highly overlapping aerial images. To evaulate the state-of-the-art the European Spatial Data Research Organisation (EuroSDR) started a benchmark on image based DSM generation. This test is based on two image blocks, of different land cover which were processed with the respective software systems. The comparison of the results provides comprehensive insight to the landscape of dense matching algorithms and demonstrates the current potential of image based photogrammetric data collection. More ...


 

For further information please contact Prof. Dr.-Ing. Norbert Haala.