In this paper, we propose a novel video denoising method applying 3D Context Model in Surfacelet Transform domain (3DCMST). Because of its directional decomposition, perfect reconstruction and low redundancy, ST has being become a powerful tool in image processing and analysis. In order to take fully advantage of the characteristic of the coefficients in ST domain, the Context model is extended from 2D to 3D, which can accomplish true 3D denoising processing. The coefficients of ST are divided into several classes according to their energy distribution by 3D Context model, and each class has independent energy estimate and threshold respectively. Experimental results show that 3DCMST has a good denoising performance in quality and fidelity. It is especially suitable for the video frame proceeding with furious movement and plenty of texture.
A Novel Video Denoising Method Based on Surfacelet Transform
2008 Congress on Image and Signal Processing ; 3 ; 245-249
2008-05-01
683438 byte
Conference paper
Electronic Resource
English
Ionograms denoising via curvelet transform
Online Contents | 2013
|EEG signal denoising based on wavelet transform
IEEE | 2017
|Coefficient denoising method with wavelet transform [3813-79]
British Library Conference Proceedings | 1999
|Continuous Wavelet Transform Denoising Method Based on Singular Value Decomposition
British Library Online Contents | 2004
|Transform domain approaches for image denoising
British Library Online Contents | 2002
|