Computer Vision deals with the process of automatic understanding of
visual information from a single-source or several images or image
sequences. From an engineering point of view it applies methods from
classical image progessing and artificial intelligence to extract
useful information from a given visual input source.
Some of the most important tasks in Computer Vision include
Based on many years of project work experience in the field of Computer vision, we would be happy to assist you as your expert contact, whether you'd like to get an overview of state-of-the-art techniques and methods or in case you have a software development request.
We worked with a variety of clients and technologies over the years of
which the most prominent are: OpenCV, ITC, VTK, Halide, HALCON,
Caffee, TensorFlow as well as C++/Qt.
Refactoring of a 100k+ codebase and performance optimization of core algorithms for a variety of target platforms and devices
Development of a software application used for segmentation and measuring structures in 3D medical images
Workshop with a focus on object detection and measurement, image segmentation, multiple-view 3D reconstruction and camera calibration
Development of a prototype for detecting inclusions and dirt in translucent gemstones
Development, optimization and parallelization of Computer Vision algorithms for the HALCON machine vision library
R&D for secure and unique detection of craquelure patterns based on image features (2D) and surface patterns (3D)
Implementation of a realtime structured-light based 3D reconstruction algorithm for an intraoral scanner
Implementation of a user friendly, easy-to-use desktop application for the 3D reconstruction of objects for a commercial CNC machine
Stereo photography on a smartphone using a prism for color channel separation and depth estimation via point correspondence
World's first mobile 3D scanner for Android, doing on-device reconstruction using Photometric-Stereo
Photometric-Stereo under unknown light source diretions using SVD
Particle filter for tracking multiple humans in high-density crowds
Implementing realtime photometric-stereo using a monochromatic camera and 8 LEDs