Institute of Information Theory and Automation

Department of Image Processing

Head of the Department:
Barbara Zitová

Deputy head of the Department:
Filip Šroubek

Jana Švarcová

phone: +420 286 581 419
staff: people, Ph.D. students
List of publications, courses, projects

The Department is involved in basic research in image processing and pictorial pattern recognition. Major application areas are biomedicine, remote sensing, astronomy, and art conservation.

Main scientific areas
  • Recognition of distorted images and patterns by invariant descriptors regardless of their actual position in the scene
  • Registration and fusion of several images of the same scene taken at different times, by different sensors and/or from different viewpoints in order to obtain information of higher quality
  • Theory of moment invariants, namely of rotation invariants, affine invariants and invariants to convolution
  • Restoration of degraded images, namely multichannel blind deconvolution, edgepreserving denoising, local contrast enhancement, and color transformations
  • Image forensics - detection of image forgeries
  • Cultural heritage applications - cooperation with art conservators in order to facilitate the conservation and material analysis work


Last events:

Jan Kotera, Ph.D. student in the Dept. of Image Processing (ZOI), received the Best Student Paper award at the Digital Photography and Mobile Imaging XI conference at IS&T/SPIE Electronic Imaging held in San Francisco, USA, for the paper "Blind Deconvolution of images with model discrepancies using maximum a posteriori estimation with heavy tailed priors", co-authored by Filip Sroubek.
The Winter School of Image Processing is a popular meeting point of students and researchers involved in image processing research. 22.1. - 26.1. 2014, Mariánská
Jan Kotera, Ph.D. student in the Dept. of Image Processing, received an award for outstanding contribution at the 15th International Conference on Computer Analysis of Images and Patterns (CAIP 2013) held in York, UK, for the paper "Blind deconvolution using alternating maximum a posteriori estimation with heavy-tailed priors" co-authored by Filip Sroubek and Peyman Milanfar.
Responsible for information: ZOI
Last modification: 22.10.2012
Institute of Information Theory and Automation