LMNO and GREYC: image denoising via block thresholding.

In many fields of application, images are omnipresent and their processing is essential if we want to make judicious use of them. Images can be subject to deterioration such as vagary perturbations, which are commonly referred to as noise. Researchers from the LMNO (CNRS, UCBN), the GREYC (CNRS, ENSICAEN, UCBN) and the CEA Saclay are now turning to a hitherto unexplored method for image denoising: the BlockJS (Block James-Stein). The BlockJS combined with recent transformations born from harmonic analysis such as curvelets, enables images to be optimally denoised with reduced calculation time.

For further information: www.cnrs.fr

Send post as PDF to