This code provides an efficient and effective color image denoising method included in our paper: Color image and multispectral image denoising using block diagonal representation
Update Version 3.0: The newest version could produce very competitive performace in terms of both efficiency and effectiveness. (9/6/2018)
Update Version 4.0: The newest implementation of our color image denoising method. (27/8/2018)
Update Version 5.0: The newest implementation of our color image denoising method. (20/9/2018)
For more information, please refer to our paper 'A Brief Review of Real-World Color Image Denoising', available at https://arxiv.org/pdf/1809.03298.pdf and also 'Color Image and Multispectral Image Denoising Using Block Diagonal Representation', available at https://arxiv.org/pdf/1902.03954.pdf
For more demos, please visit https://531623594.wixsite.com/pure-image.
@ARTICLE{Kong_TIP, author={Z. {Kong} and X. {Yang}}, journal={IEEE Transactions on Image Processing}, title={Color Image and Multispectral Image Denoising Using Block Diagonal Representation}, year={2019}, volume={28}, number={9}, pages={4247-4259}, keywords={image colour analysis;image denoising;image filtering;image representation;matrix algebra;principal component analysis;transforms;color image;multispectral image denoising;block diagonal representation;image filtering;self-similarity;sparse linear approximation;patch level representation;block diagonal matrix;global patch basis training;local principal component analysis transform;transform-threshold-inverse method;Image color analysis;Transforms;Noise reduction;Image denoising;Color;Correlation;Color image denoising;multispectral image denoising;non-local filters;transform domain techniques;block diagonal representation}, doi={10.1109/TIP.2019.2907478}, ISSN={}, month={Sep.},}