HomeNewsListProf. Thierry Blu visited on April 6, 2016

Prof. Thierry Blu visited on April 6, 2016

Prof. Thierry Blu from Chinese University of Hong Kong visited our lab on April 6, 2016. He gave a talk on "Sampling curves with Finite Rate of Innovation" in Room 9-206, Rohm Building at 2:30PM.

 

Thierry Blu was born in Orléans, France, in 1964. He received the "Diplôme d'ingénieur" from École Polytechnique, France, in 1986 and from Télécom Paris (ENST), France, in 1988. In 1996, he obtained a Ph.D in electrical engineering from ENST for a study on iterated rational filterbanks, applied to wideband audio coding. Between 1998 and 2007, he was with the Biomedical Imaging Group at the Swiss Federal Institute of Technology (EPFL) in Lausanne, Switzerland. He is now a Professor in the Department of Electronic Engineering, The Chinese University of Hong Kong. Dr. Blu was the recipient of two best paper awards from the IEEE Signal Processing Society (2003 and 2006). He is also coauthor of a paper that received a Young Author best paper award (2009) from the same society. He has been an Associate Editor for the IEEE Transactions on Image Processing (2002-2006), the IEEE Transactions on Signal Processing (2006-2010), Elsevier Signal Processing (2008-2011). He was a member of the IEEE Signal Processing Theory and Methods Technical Committee (2008-2013). He is currently on the board of Eurasip J. on Image and Video Processing (since 2010). He was elected Fellow of the IEEE in 2012 for "fundamental contributions to approximation theory in signal and image processing". Research interests: (multi)wavelets, multiresolution analysis, multirate filterbanks, interpolation, approximation and sampling theory, sparse sampling, image denoising, psychoacoustics, biomedical imaging, optics, wave propagation, etc.

 

Title:

Sampling curves with Finite Rate of Innovation

Abstract:

We extend the theory of sampling signals with "finite rate of innovation" (FRI) to a specific class of two-dimensional curves, that are defined implicitly as the zeros of a mask function. The key ingredient here is that this mask function has a low-order parametric representation and therefore, has finite rate of innovation. Then, we show that an ideal edge image whose discontinuities lie on the FRI curve satisfy a set of linear "annihilation equations" expressed in the Fourier domain. Solving these equations provides the mask function and hence, the exact edge positions in the continuous domain. Robust reconstruction algorithms are developed to deal with scenarii with model mismatch. Finally, using the fact that the annihilation equations that characterize an FRI curve are linear, we show that it is easy to include them in (convex) optimization problems. We demonstrate one potential application of the annihilation algorithm with examples in edge-preserving up-sampling. Experimental results clearly show the effectiveness of the annihilation constraint in preserving sharp edges, and improving SNR. Joint work with Hanjie Pan (EPF Lausanne, Switzerland) and Pier-Luigi Dragotti (Imperial College, London UK).