Topics for theses
Topics for Bachelor theses (BT)
- (reserviert) Evaluation of the ArcGIS Pro GeoAI Toolbox for AI-based object recognition in UAV-based aerial imagery (Contact: Melanie Elias)
- Use of high-resolution terrain models to calculate the incident solar radiation power, taking into account the local topography (in cooperation with the HZDR, contact: Anette Eltner)
- Validation of possibilities for combining a terrestrial laser scanner and a camera using the example of the Riegl VZ-400i (Contact: Katja Richter and Steffen Isfort)
Adaptable topics (suitable for BT and MT)
-
AI-based stabilization and occlusion correction of time-lapse images to optimize 4D change detection (Infos see Link, contact: Melanie Elias and Anette Eltner)
- 4D filter approaches to process time series of 3D point clouds (Contact: Melanie Elias and Anette Eltner, JProf. Geosensorsystems)
-
Analysis of the potential of the TreeAIBox for the evaluation of laser scanner data of trees in CloudCompare.
Topics for Master's theses (MT)
- (reserviert) Generation and comparison of complex 3D point clouds from dense matching and AI-generated monocular depth maps (Contact: Melanie Elias and Anette Eltner)
- Analysis of the possibilities of identifying water lines from photogrammetric data (Contact: Prof. Maas)
- Development of a method for image-based evaluation of measurements of visibility depth in water with a Secchi disk. (Contact: David Mader & Katja Richter)
- Deep learning to improve aerial photogrammetry with historical imagery. (Contact: Anette Eltner & Ferdinand Maiwald)
- Investigating Deep Learning for Stereo Matching of low-cost imagery to study river bank erosion (Contact: Melanie Elias and Anette Eltner, JProf. Geosensorsystems)
- Automatic extraction of water lines from multi-spectral aerial imagery data to derive consistent depth models in lake surveying. (in cooperation with the Bundesamt für Seeschifffahrt und Hydrographie (BSH), contact: Christian Mulsow)
- Evaluation of image processing methods for the automated detection of pedestrians in high-resolution satellite images (Contact: Melanie Elias & JProf Anette Eltner)
- Potential of AI-based software engineering for solving photogrammetric problems (programming skills required, Contact: JProf Anette Eltner und David Mader)
- Development of a method for the detection and quantitative determination of branch movements in multi-temporal laser scanner data sets (Contact: Anne Bienert and Katja Richter)
- Artificial intelligence-based classification of vegetation structures in 3D LiDAR data of forest ecosystems (Contact: Anne Bienert and Katja Richter)
Notes
Please ask the supervisors about the availability of the topics. For the selection of the topic, the processing, submission and evaluation of the thesis, please refer to our guidelines for theses.