Projects
Industrial projects since 2011 – industries, topics and technologies
My clients are covered by non-disclosure agreements, so projects are described by industry rather than by name.
Since 2011 I have worked on industrial projects in automotive production and advanced engineering projects (object detection, optical flow, SLAM, fusion with radar and lidar, calibration, stitching), in medical devices (tracking, calibration, eye tracking during eye surgery) and in industrial quality inspection (classification, detection, checks against reference images on embedded smart cameras). In addition: surveillance (video processing across several GPUs) as well as mobile and augmented reality applications.
Industries and topics
Automotive 2011–2026
Pre-development and real-time image processing: object detection with stereo, mono and fisheye cameras, lane detection, road detection and road models, optical flow, ego-motion and SLAM, environment models, occupancy grid maps and path planning, fusion with radar and lidar, online and in-field calibration, vehicle measurement and stitching. Plus test frameworks and schedulers for vision algorithms in product (embedded) and simulation.
C++ · Python · MATLAB · OpenCV · CNN · numerical optimisation · DSP · Embedded
Medical devices 2016–2025
Detection and tracking of medical devices with a stereo camera, in-field calibration of medical tools, detection of augmented reality markers, and tracking of eye movements with iris detection during eye surgery – real-time throughout and safety-relevant.
C++ · OpenCV · optical flow · Ceres · Boost · CUDA
Industrial quality inspection since 2026
Classification, detection and diff-check pipelines for embedded smart cameras that allow industrial customers to annotate, train and deploy machine vision solutions directly on the device: automated assembly verification against customer-defined reference images per work step, OK/NOK decision. Plus embedded object detection in industrial machines.
Python · PyTorch · Ultralytics/YOLO · MobileNetV3 · CNN · TensorRT · ONNX · NVIDIA Jetson · Embedded · Edge Computing
Further industries 2011–2022
Surveillance (2011): parallel decoding, rendering, image correction and stitching of multiple high-resolution video streams across several GPUs. Mobile and augmented reality (2015–2016): measuring objects and efficient packing strategies in a smartphone app, detection and realistic recolouring of wall segments, panorama stitching. Industrial software (2019–2022): debugging, development and refactoring for a manufacturer of grinding machine software.
C++ · OpenCV · CUDA · Direct3D · Swift · Objective-C · iOS · gtest
Projects from my doctorate and research work at Saarland University – together with publications, talks and reviewer activities – are listed on the research page.
From practice
Selected project examples will follow shortly.
I am happy to send a detailed profile with the full project list on request.
Get in touch
Send me a short note about the problem and we will set up a call – 30 minutes is usually enough to tell whether I can contribute something useful.
oliver@vogel-haus.de · +49 172 961 87 23
