Teaching

If you are interested in writing your thesis in the Computer Vision in Engineering group, we offer a range of courses that provide a solid foundation in relevant subject areas. We recommend that you first take appropriate courses and then complete a Hot Topics in Computer Vision (HTCV) project to gain in-depth insight into specific areas.
After completing an HTCV project, there is usually an opportunity to discuss potential topics for a thesis. Supervision is mainly provided by research assistants or PhD students in the group working on topics related to their current research.
In order to identify potential thesis topics, we recommend that you check our current research projects and the profiles of our team members. The first reviewer for all theses will be Prof. Rodehorst, the second reviewer can be another professor, the supervisor or an external reviewer, depending on the study program.
If you have any further questions, please contact cv.thesis[at]medien.uni-weimar.de.

Master's theses

• Narayana Manikanta Sunnit Saty Gattupalli Eswar: Material-based Segmentation in Texture Space, SoSe 2026
• Ganesh Kumar Uppiretla: Comparative Analysis of Calibrated and Uncalibrated Photometric Stereo with Area and Point Light Sources, SoSe 2026
• Sahand Majdabadi Farahani: Reproducing and Evaluating State-of-the-Art Methods in Chest X-ray Abnormalities Detection, WiSe 2025/26
• Adrian Teschendorf: Reconstruction of 3D Faces with Animatable Facial Expressions, WiSe 2024/25
• Naga Padma Veerla: A Gaussian Splatting Framework for 3D Object Recognition and Segmentation in Agriculture, WiSe 2024/25
• Lucky Chandrautama: Consistency Metrics for Text-Based Stylization in Multi-View Images of 3D Gaussian Splatting Head Avatars, WiSe 2024/25
• Philipp Tornow: Open-Vocabulary Semantic Segmentation of 3D Gaussian Splatting Radiance Fields, SoSe 2024
• Ella Maria Gekeler: Construction Industry Steel Ordering Lists (CISOL) Dataset - A data driven approach to table detection and table structure recognition, SuSe 2024
• Henrik Norderhus: CrackSense: A Pruned Neural Network Approach for Crack Detection on Embedded Devices, WiSe 2023/24
• Morris Florek: Efficient and Discriminative Image Feature Extraction for Multi-Domain Image Retrieval, WiSe 2023/24
• Raghu Varma Yarakaraju: Finding Image-to-Image Similarities in a Historic Humanities Corpus, SoSe 2022
• Julian Steinhaußen: Deep Learning for Image Registration, SuSe 2022
• El Mehdi Essadki: A Framework for Guided 3D Point Cloud Annotation, WiSe 2021/22
• Benjamin Burse: A modular approach towards cost-efficient monitoring based on microcontrollers, SuSe 2021
• Pia Fichtl: Extraction of roof shapes for city models from UAS, SuSe 2021
• Huy Khanh Ha: Real-time multi-view marker-less motion capture of multiple humans, SuSe 2021
• Jan Frederick Eick: Deep learning-based stereo matching for high-resolution images, WiSe 2020/21
• Kai Gerrit Lünsdorf: Deep learning-based camera pose regression for bridge inspection, WiSe 2020/21
• David Tschirschwitz: Deep learning for semantic segmentation – using CycleGAN to detect structural defects with sparse and unlabeled data, WiSe 2020/21
• Srividya Raju: Fusion of convolutional neural network results from pixel-wise image segmentation, WiSe 2020/21
• Michael Ohaya: Classification of crack and non-crack images using convolutional neural networks, SuSe 2019
• Paul Debus: Automatic pathplanning in UAS building survey, SuSe 2018
• Darshan Patel: Real-time obstacle detection and collision avoidance for autonomous UAS navigation, WiSe 2017/18
• Amir Azimian: Image analysis of existing building as a basis for the simulation of virtual cities, SoSe 2017
• Endre Adalar Papp: Unmanned aerial system framework towards semi-autonomous flight and indoor mapping, SuSe 2017
• Asma Naz: Analysis and implementation of image enhancement approaches for 3D reconstruction, WiSe 2016/17
• Felicitas Höbelt: Cross-scale cost aggregation and PatchMatch filter: an improved dense stereo matching method, WiSe 2016/17
• Jingxi Xu: Single image super-resolution via self similarity exploration and steering kernel regression, SuSe 2016
• Tejeswini Kashyappan: Convolutional neural networks for object detection and recognition, SuSe 2016
• Bastian Weber: Implementation and parallelisation of cone-beam computed tomography backprojection on CPU and GPU, SuSe 2016
• Suvratha Narayan Bejai: Automated marker detection and identification, SuSe 2016
• Rashmi Rangaswamaiah: Optical flow for visual odometry, SuSe 2016
• Jenny Carolina Gonzalez Acuna: Real-time object tracking in videos, SuSe 2016
• A. Hamid Sabiri: Image forgery detection – detecting duplicated regions in an image, SuSe 2016
• Pranoy Vijaykumar: Knowledge-based pattern recognition method for crack detection, SuSe 2016
• Markus Kossatz: Visualizing and editing the history of countries in time and space with HistoGlobe, SuSe 2016
• Malik Al-hallak: A vision-based lane detection and tracking system, SuSe 2016
• Martin Horbach: Real-time stereo matching on GPU, SuSe 2016
• Steve Herzog: Quality assurance in structural steel engineering using terrestrial laser scanning, WiSe 2015/16
• Jenis Jetani: Preprocessing methods for CT images, SuSe 2015
• Pavlos Iliopoulos: Automated depth from focus with a DSLR camera, SuSe 2015
• Christian Baar: 3D motion analysis of athletes in real conditions, SuSe 2015
• Bastian Karge: Structure-from-Motion: 3D reconstruction on mobile devices, WiSe 2014/15

Bachelor's theses

• Aaron Kammer: Geometry Informed Optimization for 3D Gaussian Splatting in Indoor Scenes, WiSe 2026
• Judith Klix: Investigation of the influence of image quality on the detection of surface defects on concrete, SoSe 2024
• Falko Horn: Combination of photogrammetry and laser scanning for neural radiation fields, WiSe 2023/24
• Florian Dietz: Comparison of acquisition techniques for the geometric as-build survey of an existing highway bridge, WiSe 2022/23
• Martin Tippmann: Retrieval of 3D Point Cloud Submaps from 2D Query Images, WiSe 2022/23
• Gissu Valentina Naghavi: Geometric calibration of RGB and thermal imaging cameras to generate a multispectral data set, SuSe 2022
• Klaus Kraemer: Development of an approach for the determination of window areas in urban district models using image recognition methods, WiSe 2021/22
• Jonas Linß: Drone Flight Path Planning - Viewpoint Calculation With UV Texture Unwrapping, WiSe 2021/22 
• Marie Lücke: Measurement data based evaluation of structures susceptible to stress corrosion cracking, SuSe 2021
• Nhu Quang Dang: Methods of geometric light source calibration, SuSe 2021
• Philipp Tornow: Unsupervised object extraction for semantic segmentation of bridge scenes, SuSe 2021
• Charlotte Marx: Comparison of acquisition methods in historic preservation inventory documentation, SuSe 2020
• Theresa Wenig: Determination of component deformations with optical flow methods, SuSe 2020
• Florian Madeya: Image-based 3D reconstruction as an animation method, WiSe 2017/18
• Frederic Gaillard: Sensor fusion with extended Kalman filter for UAS navigation, SuSe 2017
• Stefan Müller: Possibilities of virtual reconstruction, SuSe 2016
• Kai Gerrit Lünsdorf: On the accuracy of the Hokuyo UST-20LX for use on unmanned aerial systems, SuSe 2016
• Hagen Hiller: Stereo-based real-time obstacle detection for unmanned aerial systems, WiSe 2015/16
• Veronika Haaf: Image analysis for detection and measurement of concrete cracks, SuSe 2015
• Adrian Teschendorf: Analysis of influencing factors for image-based 3D reconstruction, SuSe 2015
• Mischa Linus Krempel: Design and implementation of memory-efficient stereo matching algorithms, SoSe 2014
• Julius Bullinger: Building monitoring with networked heterogeneous sensor architectures, SuSe 2014
• Stefanie Riegg: Investigation of deformations of planar and irregular point clouds, SuSe 2014
• Eugen Müller: Automatic image analysis for crack detection on concrete surfaces, SuSe 2014
• Andreas Aicher: Comparative analysis of measurement series of different measurement systems for the determination of vertical structural deformations for structural safety evaluations, 2014
• Michael Wachtmann: Investigations on the quality of freely available geodata, WiSe 2013/14
• Max Patzelt: Investigations into the performance potential of close-range scanners using reconstruction as an example, 2013