Home

Awesome

The code in this package implements the Trilateral Weighted Sparse Coding Scheme for real color image denoising as described in the following paper:

@article{TWSC_ECCV2018,        
         author = {Jun Xu and Lei Zhang and David Zhang},        
         title = {A Trilateral Weighted Sparse Coding Scheme for Real-World Image Denoising},        
         journal = {ECCV},       
         year = {2018}     
}

Please cite the paper if you feel this code useful in your research. Please see the file License.txt for the license governing this code.

Version: 1.0 (13/07/2018), see ChangeLog.txt Contact: Jun Xu <csjunxu@comp.polyu.edu.hk, nankaimathxujun@gmail.com>

Test

  1. Run "Demo_TWSC_Sigma_AWGN.m" for Additive White Gaussian noise removal.
  2. Run "Demo_TWSC_Sigma_RW*.m" for Real-world noise removal. Note: Please set "Original_image_dir" according to your case.

Data

Please download the data from corresponding addresses.

  1. cleanimages: 20 high quality commonly used natural gray scale images

  2. nc: real noisy images with no ''ground truth'' This dataset can be found at http://demo.ipol.im/demo/125/

  3. cc: 15 cropped real noisy images from CC [1]. This dataset can be found at http://snam.ml/research/ccnoise The smaller 15 cropped images can be found on in the directory ''Real_ccnoise_denoised_part'' of https://github.com/csjunxu/MCWNNM_ICCV2017 The *real.png are noisy images; The *mean.png are "ground truth" images; The *ours.png are images denoised by CC.

  4. dnd: The Darmstadt Noise Dataset [2] consists of 50 pairs of real noisy images, each images provides 50 crops, resulting overall 1,000 crops provided on https://noise.visinf.tu-darmstadt.de/

[1] A Holistic Approach to Cross-Channel Image Noise Modeling and its Application to Image Denoising. Seonghyeon Nam*, Youngbae Hwang*, Yasuyuki Matsushita, Seon Joo Kim. CVPR 2016.

[2] Benchmarking Denoising Algorithms with Real Photographs. Tobias Plötz and Stefan Roth. CVPR 2017.

Dependency

This code is implemented purely in Matlab2014b and doesn't depends on any other toolbox.

Contact

If you have any questions or suggestions with the code, or find a bug, please let us know. Contact Jun Xu at csjunxu@comp.polyu.edu.hk or nankaimathxujun@gmail.com.