At PTS Heidenau, I prototyped a Python-based tool to bring scanner-calibration and image-processing steps into a repeatable workflow. I processed IT8 target data and TIFF images, and evaluated OCR/EasyOCR methods for detecting text near colour-patch regions.
At TU Dresden, I'm developing a workflow that aligns CAD data with high-resolution X-ray PCB images and generates YOLO-format labels. I also built and evaluated a speech-driven robotic control workflow using Python, PyTorch, ROS 2, Whisper, and NVIDIA ASR.

