Themen Abschlussarbeiten
- Machen Sie in Ihrer Anfrage kenntlich an welchen konkreten Themen Sie interessiert sind und wieso Sie sich hierfür interessieren. Die verfügbaren Themengebiete finden Sie unten. Gerne können Sie ihre Anfrage auch direkt an die entsprechenden Mitarbeiter:innen senden.
- Machen Sie sich mit unserer aktuellen Forschung in den für Sie interessanten Bereichen vertraut. Aktuelle Projektinformationen sowie Publikationen finden Sie auf den Seiten der jeweiligen Mitarbeiter:innen.
- Fügen Sie Ihrer Anfrage ein aktuelles Transcript of Records sowie eine kurze Beschreibung ihrer relevanten Kenntnisse und Fähigkeiten bei. Beachten Sie hierbei auch die notwendigen Anforderungen für die Bearbeitung der Abschlussarbeit, diese finden Sie innerhalb der jeweiligen Ausschreibungen. Anfragen ohne die oben genannten Informationen können leider nicht beantwortet werden.
Abschlussarbeiten
Verfügbare Themen für Abschlussarbeiten
Abschlussarbeiten in Bearbeitung
Semi-Supervised Bubble Segmentation for Video Capsule Endoscopy
Abstract
Video capsule endoscopy produces vast numbers of frames per examination, particularly of the small bowel, which is difficult to reach by other means, and manually reviewing and labelling this material is impractical. This thesis therefore investigates bubble segmentation in capsule endoscopy frames using semi-supervised learning, so that a small labelled subset suffices to classify the entire dataset. A Python/OpenCV tool is developed to annotate frames and generate bitmasks marking bubble positions; these masks then train a U-Net-based deep CNN in PyTorch, pre-trained on RGB data, whose predictions are evaluated using accuracy and precision scores.
Integration of a Hardware Accelerator for On-Device Localization in an FPGA-Based Video Capsule Prototype
Abstract
In video capsule endoscopy, most of the capsule’s limited energy budget is spent on illumination and the wireless transmission of every captured image, even though many frames show organs irrelevant to the examination. This thesis integrates the resource-efficient UltraTrail AI accelerator into the FPGA of an Ovesco capsule prototype, using a Lattice CrossLink-NX-33 and a SystemVerilog design, so that a compact neural network can localize the capsule within the GI tract on-device and restrict transmission to the organ of interest. The design is verified in simulation and tested on real hardware via a USB-to-SPI adapter and a hardware-in-the-loop camera model, and is extended with improved memory organization, a Hidden Markov Model for more accurate localization, and pipelining of image capture with model execution to further reduce energy consumption.Abgeschlossene Abschlussarbeiten
Evaluation of Energy-Efficient Image Compression Algorithms for Video Capsule Endoscopy