Learning how to combine internal and external denoising methods

Harold Christopher Burger

Christian J. Schuler

Stefan Harmeling


Different methods for image denoising have complementary strengths and can be combined to improve image denoising performance, as has been noted by several authors [11, 7]. Mosseri et al. [11] distinguish between internal and external methods depending whether they exploit internal or external statistics [13]. They also propose a rule-based scheme (PatchSNR) to combine these two classes of algorithms. In this paper, we test the underlying assumptions and show that many images might not be easily split into regions where internal methods or external meth- ods are preferable. Instead we propose a learning based approach using a neural network, that automatically combines denoising results from an internal and from an external method. This approach outperforms both other combination methods and state-of-the-art stand-alone image denoising methods, hereby further closing the gap to the theoretically achievable performance limits of denoising [9]. Our denoising results can be replicated with a publicly available toolbox.

Paper (pdf) 2MB
Supplementary material (pdf) 24MB


This Matlab toolbox combines the results of the toolbox above and of BM3D, using neural networks. For more details, see Burger's PhD thesis. This method achieves the best denoising results reported in literature by any method.
Matlab toolbox (zip)
Neural networks for various noise levels:
sigma=10 sigma=25 sigma=35 sigma=50 sigma=75 sigma=170 Warning: Large files!