Artifacts suppression in images and video Volodymyr Fedak
- Slides: 13
Artifacts suppression in images and video Volodymyr Fedak
Introduction l What is the problem? l Why is it important? l What did I do? l What are the results? l So what next?
What is the problem? l blocking l ringing l blurring l flickering
What is the problem? F-2 F-1 F F+1 F+2 Intra-frame processing… Inter-frame processing…
Why is it important ? Postprocessing techniques: • spatial-temporal algorithms • algorithms that transform signal to frequency domain Compressed information De-coder Artifact detection Coder parameters • motion-compensated algorithms • iterative approaches based on theory of projections onto convex set Reducing artifacts postprocessing Transform to original format Enhanced information
What did I do ? l Analyse modern postprocessing techniques l Implement most encouraging methods l Compare results of mentioned algorithms l Propose approaches for optimization
Wavelet-based de-blocking and de-ringing algorithm proposed by Alan and Liew Steps: • Detection of Block Discontinuities • Threshold Maps Generation at Different Wavelet Scales • low frequency filtering
Non-Local Means NLM is an improvement of Bilateral filtering C(y, x) - geometric relationship S(I(y), I(x)) - luminance ratio I(y) – pixel luminance
Non-Local Means NLM could be presented: in general way: in terms of implementation: v(i) – noisy image W(i, j) - weighted average of pixels in the image v(j) – pixel luminance N(x) - window surrounding pixel x; Q(x) is a search window around pixel x;
Non-Local Means Parameters • h - determines the amount of averaging (h increases amount of blocking artifacts decrease). • N (x) – the match window/patch – when N(x) increases, blocking artifacts of the processed sequence decreases very slowly • Q(x) – the search window/patch – when Q(x) increases, artifacts of the processed sequence decreases very slowly for an increasing value of the search window size, and we have a large amount of computation time.
Possible ways for optimization: • Extended NLM to the temporal domain. Use together with motion-compensation algorithm but apply some quality coefficient to the motion vector. • Add smart patch/search window size choosing algorithm. • Use Hierarchical block matching algorithm to find similar windows for speeding-up NLM
Any questions ?
