Abstract

Optimal denoising works at best on raw images (the image formed at the output of the focal plane, at the CCD or CMOS detector), which display a white signal-dependent noise. The noise model of the raw image is characterized by a function that given the intensity of a pixel in the noisy image returns the corresponding standard deviation; the plot of this function is the noise curve. This paper develops a nonparametric approach estimating the noise curve directly from a single raw image. An extensive cross-validation procedure is described to compare this new method with state-of-the-art parametric methods and with laboratory calibration methods giving a reliable ground truth, even for nonlinear detectors.

© 2014 Optical Society of America

Gaussian noise level estimation for color image denoising

Xue Guo, Feng Liu, and Xuetao Tian
J. Opt. Soc. Am. A 38(8) 1150-1159 (2021)

Adaptive noise estimation from highly textured hyperspectral images

Peng Fu, Changyang Li, Yong Xia, Zexuan Ji, Quansen Sun, Weidong Cai, and David Dagan Feng
Appl. Opt. 53(30) 7059-7071 (2014)

Noise-insensitive image optimal flow estimation using higher-order statistics

El Mehdi Ismaili Alaoui, Elhassane Ibn-elhaj, and El Houssaine Bouyakhf
J. Opt. Soc. Am. A 26(5) 1212-1220 (2009)

Full-spectrum denoising of high-SNR hyperspectral images

Miguel Colom and Jean-Michel Morel
J. Opt. Soc. Am. A 36(3) 450-463 (2019)

Minimum variance unbiased subpixel centroid estimation of point image limited by photon shot noise

Hui Jia, Jiankun Yang, and Xiujian Li
J. Opt. Soc. Am. A 27(9) 2038-2045 (2010)