Removal of noise from video

Contact person: Dmitriy S. Vatolin (dmitriy@graphics.cs.msu.ru)

MSU Graphics & Media Lab (Video Group)

The filter is designed for noisy video stream processing, video quality improvement, or video preprocessing (e.g. for preprocessing before compression or deinterlacing). The algorithm preserves picture details while performing noise suppression.

Examples

Let's consider the filter operation on the Susy sequence first. The sequence is quite noisy – spatial noise is easy to observe, and temporal noise is computed and visualised using LUV metric.

Source frame with noise LUV metric for frames 13 and 14

There are plenty of green dots in LUV metric, telling about the presence of temporal noise. Red spots show where motion takes place.

Let's compare source frame with the frame processed by MSU Denoising Filter:

Source frame with noise Frame processed by MSU Denoising Filter
(made with 'hard' preset that emphasizes filter effect)


LUV metric for source and processed frames. LUV metric tells that we managed to get rid of the most of spatial noise

Fragments of the same frame are shown below.

Source frame with noise Frame processed by MSU Denoising Filter
(made with 'hard' preset that emphasizes filter effect)

And another fragment of the same frame:

Source frame with noise Frame processed by MSU Denoising Filter
(made with 'hard' preset that emphasizes filter effect)

Finally, consider MSU Denoising Filter application to Arm sequence from Terminator 2 movie.

Source frame with noise Frame processed by MSU Denoising Filter
(made with 'hard' preset that emphasizes filter effect)

Download

Filter for VirtualDub MSU Denoising (79Kb, ZIP)
Unpack this file in the folder "Plugins" in VirtualDub directory to use.

See also

Detailed description of MSU Denoising filter
Description and comparison of noise suppression methods used in other available filters.
Description of MSU Denoising (free) & MSU Noise Removal filters.
In this report we show how MSU Noise Removal usage can increase compression up to 30% with better visual quality.

Team