Image and Video Recovery

Imaging systems are ubiquitous and indispensable in science and everyday life. Examples range from astronomy to commercial photography and medical diagnostic systems, like magnetic resonance (MR) and computer-assisted tomography (CT), to fluorescence microscopy, to name a few. A common problem with all imaging systems is that they can only acquire a degraded version of the original scene. The image recovery problem is then the problem of uncovering with digital signal processing approaches the information lost due to any of the reasons mentioned above, using the available data. Expressed differently, an estimate of the original image is sought utilizing the available data and prior knowledge about the degradation system and the original image. Specific problems in image recovery are restoration, blind deconvolution, denoising, super-resolution, deblocking, concealment and inpainting problems.

IVPL has been very active in nearly all of these topics and made significant contributions on both theoretic and algorithmic levels. You can check out the relevant publications in this area listed below. More information on recent projects we are working on can be found below.