Dominic Rüfenacht

Image and Video Representation Group (IVRG)
School of Computer and Communication Sciences (IC)
Ecole Polytechnique Fédérale de Lausanne (EPFL)

Email: dominic [dot] ruefenacht [at] a3 [dot] epfl [dot] ch

I have left EPFL end of February 2013 to start a PhD at the University of New South Wales in Sydney, Australia. You can still reach me using the following email address:

dominic[dot]ruefenacht [at] a3 [dot] epfl [dot] ch

Research Interests

  • Computational Photography
    – High Dynamic Range (HDR) Imaging
    – Near Infrared (NIR) Imaging
  • Camera Processing Pipeline


Automatic and Accurate Shadow Detection using Near-Infrared Information

  We present a method to automatically detect shadows in a fast and accurate manner by taking advantage of the inherent sensitivity of digital camera sensors to the near-infrared (NIR) part of the spectrum. Dark objects, which confound many shadow detection algorithms, often have much higher reflectance in the NIR. We can thus build an accurate shadow candidate map based on image pixels that are dark both in the visible and NIR representations. We further refine the shadow map by incorporating ratios of the visible to the NIR image, based on the observation that commonly encountered light sources have very distinct spectra in the NIR band. The results are validated on a new database, which contains visible/NIR images for a large variety of real-world shadow creating illuminant conditions, as well as manually labelled shadow ground truth. Both quantitative and qualitative evaluations show that our method outperforms current state-of-the-art shadow detection algorithms in terms of accuracy and computational efficiency.


Temporally Consistent Snow Cover Estimation from Noisy, irregularly sampled Measurements

We propose a method for accurate and temporally consistent surface classification in the presence of sparse, noisy and irregularly sampled measurements, and apply it to the estimation of snow coverage over time. The input imagery is extremely challenging, with large variations in lighting and weather distorting the measurements. Initial snow cover estimations are obtained using a Gaussian Mixture Model of color. To achieve a temporally consistent snow-cover estimation, we use a Markov Random Field that penalizes rapid fluctuations in the snow state, and show that the penalty term needs to be quite large, resulting in slow reactivity to changes. We thus propose a classifier to separate good from uninformative images, which allows to use a smaller penalty term. We show that the incorporation of domain knowledge to discard uninformative images leads to better reactivity to changes in snow coverage as well as more accurate snow cover estimations.


Stereoscopic High Dynamic Range Imaging
Master Thesis, Philips, Netherlands (Feb. – Aug. 2011)

   The thesis investigated the possibility of recording stereoscopic HDR video using LDR cameras. On one side, traditional stereo setups require the two cameras to have the same intrinsic parameters, and to take images from slightly different points of view. This means that the exposure time is the same for the two cameras. On the other side, HDR imaging needs the exposures to capture different parts of the dynamic range by changing the exposure time of the captures, as well as having those exposures aligned. By combining stereo with HDR, we need to violate those fundamental assumptions. In other words, we try to combine two domains that seem to be mutually exclusive. We investigate two different modes to record stereoscopic HDR video, which in turns influence the way the frames are processed.



  • 2009-2011: M.Sc. in Communication Systems, EPFL, Switzerland
  • 2008-2009: Exchange year at University of Waterloo, Canada
  • 2005-2009: B.Sc. in Communication Systems, EPFL, Switzerland

Working Experience

  • Sept. 2011 – Feb. 2013: Scientific Assistant at the IVRG, EPFL, Switzerland
  • Feb. – Aug. 2011: Master Thesis in Industry: Stereoscopic High Dynamic Range Imaging, Philips, Netherlands
  • Spring Semester 2010: Teaching Assistant for course “Digital Photography”, EPFL, Switzerland



D. Rüfenacht; M. Brown; J. Beutel; S. Süsstrunk : Temporally Consistent Snow Cover Estimation from Noisy, Irregularly Sampled Measurements. 2014. 9th International Conference on Computer Vision Theory and Applications (VISAPP), Lisbon, Portugal, January 5-8, 2014.
D. Rüfenacht; C. Fredembach; S. Süsstrunk : Automatic and Accurate Shadow Detection using Near-Infrared Information; IEEE Transactions on Pattern Analysis and Machine Intelligence. 2014. DOI : 10.1109/TPAMI.2013.229.


D. Rüfenacht; G. Trumpy; R. Gschwind; S. Süsstrunk : Automatic Detection of Dust and Scratches in Silver Halide Film using Polarized Dark-Field Illumination. 2013. IEEE 20th International Conference on Image Processing (ICIP), Melbourne, Australia, September 15-18, 2013. p. 2096-2100. DOI : 10.1109/ICIP.2013.6738432.


C. Fredembach; D. Rüfenacht; S. Süsstrunk : Automatic and Accurate Shadow Detection using Near-Infrared Information ; IEEE International Conference on Computational Photography, Seattle, WA, USA, April 27-29, 2012.


D. Rüfenacht : Stereoscopic High Dynamic Range Video ; 2011.


D. RUFENACHT; R. DUMUSC; F. DUTOIT; M. STAEHLI; M. HALLER : Aventures en Suisse Romande ; 2010.