Regularization by Denoising uses a denoising engine to solve many image reconstruction ill-posed inverse problems. This is a R implementation of the algorithm developed by Romano et.al. (2016) <arXiv:1611.02862>. Currently, only the gradient descent optimization framework is implemented. Also, only the median filter is implemented as a denoiser engine. However, (almost) any denoiser engine can be plugged in. There are currently available 3 reconstruction tasks: denoise, deblur and super-resolution. And again, any other task can be easily plugged into the main function 'RED'.
| Version: | 1.0.0 |
| Depends: | R (≥ 3.4.0), imager |
| Published: | 2017-06-14 |
| Author: | person("Adriano", "Passos", email="adriano.utfpr@gmail.com", role=c("aut","cre")) |
| Maintainer: | Adriano G. Passos <adriano.utfpr at gmail.com> |
| License: | GPL-3 |
| NeedsCompilation: | no |
| CRAN checks: | redR results |
| Reference manual: | redR.pdf |
| Package source: | redR_1.0.0.tar.gz |
| Windows binaries: | r-devel: redR_1.0.0.zip, r-release: redR_1.0.0.zip, r-oldrel: redR_1.0.0.zip |
| OS X binaries: | r-release: redR_1.0.0.tgz, r-oldrel: redR_1.0.0.tgz |
Please use the canonical form https://CRAN.R-project.org/package=redR to link to this page.