Markus Blog

About me

Computer Vision, Signal Processing, Sensing, ML, Data & Signal Analysis

I have been developing algorithms and working on computer vision systems for over two decades. At universities, large companies, startup, and freelance.

From hardware, optics, and system integration, (multi-) camera models and calibration. On to compression, image signal and 3D reconstruction, signal processing. Lastly machine learning, and deep learning from training data management to release management.

Requirements Engineering, iOS App Development

I have developed the technical architecture for two early startups. With a close eye on performance, user experience, data security, product development.

Past Responsibilities

  • Technical Architect in early Startups
  • Algorithm and DNN Development (mostly C++ and Python)
  • Mobile Computer Vision SDK Development (iOS, Android, C++, Python, Cordova, .NET MAUI)
  • ML Deployment Pipeline for Mobile CV SDK
  • Mobile App development, mostly iOS, some Android, .NET
  • UI and UX concepts (focussing on camera based UIs)
  • Various algorithmic, runtime and data handling contributions in a big data Python customer project
  • Project Management & Controlling
  • Terms of use, Privacy policy, General terms and conditions documents
  • Publicly Funded Projects: proposal, management, controlling, reporting
  • Patents and Norms
  • Student Supervision (Master and Bachelor Theses, Practical Semesters)

Computer Vision Experience

  • DNN-based Scan Module for analog and digital Meters
  • Scan Module for Gauges
  • Scan Module for Text, e.g. Serial Numbers, IBANs, Health Monitoring Devices, etc.
  • Scan Module for Paint Coverage (Quality Control)
  • Sensor Fusion: Camera + Accelerometer, Camera + Laser Range
  • Extrinsic, Intrinsic, and Tonal Camera Calibration
  • Multi-Camera Exposure Alignment
  • Restoration Filtering of Bayer-Pattern-based RGB Images
  • Restoration Filtering of Compressed Videos or Images
  • Temporal Video Compression Optimization
  • Video Compression at Multiple Resolutions

Publicly Funded Projects

  • BSFZ Siegel for DNN-Upgrade of pixometer SDK
  • eVIDENCE - Probative Scan of Meter Values
  • KOMEET - Mobile Documentation in Intensive Care
  • V-Charge - Autonomous Valet Parking and Charging for e-Mobility

Patent Applications

  • EP4187217A1 - Mobile Rechnervorrichtung zur Durchführung von Farbechtheitsmessungen
  • DE102016119071A1 - Bilderfassung
  • DE102013211752A1 - Werkzeugsystem mit einer handgeführten Werkzeugmaschine
  • DE102010042693A1 - Verfahren zur Aufbereitung eines Bilddatenstroms

Past Affiliations

  • RWTH Aachen University
  • University of Ottawa
  • Bosch Sicherheitssysteme
  • Bosch Corporate Research
  • pixolus GmbH
  • freelance