Research topics and projects

Our research group has a keen interest in all aspects of image and video processing. We would like to continue the close collaboration that we have now with the other image processing groups at Ghent University and abroad. We are always willing to guide outsiders (e.g., industrial representatives) into the domain of image processing or to help them solve problems related to image processing (e.g., by defining graduate thesis work, or by joint project applications, or through consulting).

Our topics can be divided in four categories:

  • Autonomous intelligent vehicles

    Increasing traffic density as a result of urbanization is posing serious challenges for our mobility infrastructure. In order to maintain acceptable levels of efficiency and safety in this increasingly complex traffic situation, intelligent vehicles support the driver with extra information or by automating parts of the driving tasks. Below is a selection of research related to this topic.

  • GPU-accelerated image/video processing-Quasar

    To reach a good computational performance, modern GPUs bring speedups of 10x-200x for highly parallel processing tasks, but one main disadvantage is the difficulty of programming: not only does (properly) programming a GPU require an extensive in-depth knowledge of the details of a GPU. Using Quasar, a high-level language like Matlab/Python, GPU/CPU programming can be done straightforwardly in a hardware agnostic manner.

  • Image/video quality assessment

    This covers our work on quality assessment of broadcast, surveillance, multimedia and medical images/videos/volumes, to evaluate technologies such as color correction, camera automation, video denoising and video compression.

  • Image/video quality enhancement

    This covers our work on quality enhancement of non-medical images and video sequences, including but not limited to: denoising depth images from time-of-flight cameras, superresolution, interpolation, demosaicing, error concealment, denoising of general images,...

  • Intelligent surveillance

    This covers our work on analysis of surveillance video sequences, including but not limited to: combining information in camera networks, traffic video analysis, foreground-background segmentation etc.

  • Medical and biological image processing

    This covers our work on medical images, including but not limited to: MRI reconstruction through compressive sensing, denoising, segmentation and analysis of medical images, ...

  • Remote sensing

    This covers our work on hyperspectral images, including but not limited to: hyperspectral image fusion, classification, segmentation and spatial analysis, ...

  • Van Eyck Art work restoration and analysis

    This work covers research of techniques to process digitized artwork images, like the Ghent Altarpiece (aka Adoration of the Mystic Lamb). Techniques include digital restoration of cracks in the painting and analysis of painted pearls.