Learn R Programming

redR (version 1.0.1)

REgularization by Denoising (RED)

Description

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) . 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'.

Copy Link

Version

Install

install.packages('redR')

Monthly Downloads

26

Version

1.0.1

License

GPL-3

Maintainer

Adriano Passos

Last Published

September 3rd, 2018

Functions in redR (1.0.1)

resample

Resampling of an image
shift

shifting operator
RED

RED: Regularization by Denoising
degrade

Degradation of an image
transform

Transform an image
error

Error measurements of images
fft_convolve

Convolution of two images via FFT
lenna

Photograph of Lenna
cameraman

Photograph of a cameraman
register

Registration parameter estimation