Second Order Differences for Combined Cyclic and Vector Space Data and their Application to Denoising and Inpainting using a TV-type Approach

Add to your list(s) Download to your calendar using vCal

• Ronny Bergmann
• Wednesday 04 March 2015, 15:00-16:00
• MR 14, Centre for Mathematical Sciences.

If you have a question about this talk, please contact Dr Jan Lellmann.

In many image applications, we have to deal with noisy measurements or even lossy data. In many cases, image data is vectorial, e.g. for color images, or even a combination of cyclic and linear space data, like for the HSV color space. In order to restore real valued data, TV minimization is quite popular. Extending this to higher order differences avoids the so called `stair casing` effect.

This talk introduce second order differences for combined cyclic and linear space data, like HSV . We present a second order total variation type functional for the inpainting and the denoising situation and the combination of both. We further derive and efficient algorithm to minimize this functional and apply the algorithms to concrete problems.

This talk is part of the Cambridge Image Analysis Seminars series.

This talk is included in these lists:

• Cambridge Centre for Data-Driven Discovery (C2D3)
• Cambridge Image Analysis Seminars
• Cambridge talks
• Chris Davis' list
• Interested Talks
• MR 14, Centre for Mathematical Sciences
• Trust & Technology Initiative - interesting events
• bld31
• ndk22's list
• ob366-ai4er
• rp587

Note that ex-directory lists are not shown.