TID - Targeted Image Denoising

Adaptive image denoising by targeted databases

This package provides an implementation of an adaptive image denoising algorithm using targeted databases. The proposed method [1, 2], called Targeted Image Denoising (TID), applies a group sparsity minimization and a localized prior to learn the optimal denoising filter from the targeted database. To have an overall evaluation of the denoising performance, please run the demo file: "demo.m". For comparison purposes, we also provide the codes for some state-of-the-art denoising methods including BM3D, BM3D-PCA, LPG-PCA, and NLM. All these methods are re-implemented and modified by us such that patch search is performed over the targeted external databases.
For additional information and citations, please refer to:
[1] E. Luo, S. H. Chan, and T. Q. Nguyen, "Adaptive Image Denoising by Targeted Databases," IEEE Trans. Image Process., vol. 24, no. 7, pp. 2167-2181, Jul. 2015.
[2] E. Luo, S. H. Chan, and T. Q. Nguyen, "Image Denoising by Targeted External Databases," in Proc. IEEE Intl. Conf. on Acoustics, Speech and Signal Process.(ICASSP'14), pp. 2469-2473, May 2014.