TUTCRIS - Tampereen teknillinen yliopisto

TUTCRIS

Unbiased Injection of Signal-Dependent Noise in Variance-Stabilized Range

Tutkimustuotosvertaisarvioitu

Standard

Unbiased Injection of Signal-Dependent Noise in Variance-Stabilized Range. / Borges, Lucas; Vieira, Marcelo; Foi, Alessandro.

julkaisussa: IEEE Signal Processing Letters, Vuosikerta 23, Nro 10, 2016, s. 1494-1498.

Tutkimustuotosvertaisarvioitu

Harvard

Borges, L, Vieira, M & Foi, A 2016, 'Unbiased Injection of Signal-Dependent Noise in Variance-Stabilized Range', IEEE Signal Processing Letters, Vuosikerta. 23, Nro 10, Sivut 1494-1498. https://doi.org/10.1109/LSP.2016.2601689

APA

Borges, L., Vieira, M., & Foi, A. (2016). Unbiased Injection of Signal-Dependent Noise in Variance-Stabilized Range. IEEE Signal Processing Letters, 23(10), 1494-1498. https://doi.org/10.1109/LSP.2016.2601689

Vancouver

Author

Borges, Lucas ; Vieira, Marcelo ; Foi, Alessandro. / Unbiased Injection of Signal-Dependent Noise in Variance-Stabilized Range. Julkaisussa: IEEE Signal Processing Letters. 2016 ; Vuosikerta 23, Nro 10. Sivut 1494-1498.

Bibtex - Lataa

@article{71d8ee454b334e59bb22180b0237948d,
title = "Unbiased Injection of Signal-Dependent Noise in Variance-Stabilized Range",
abstract = "The design, optimization, and validation of many image processing or image-based analysis systems often requires testing of the system performance over a dataset of images corrupted by noise at different signal-to-noise ratio regimes. A noise-free ground-truth image may not be available, and different SNRs are simulated by injecting extra noise into an already noisy image. However, noise in real-world systems is typically signal-dependent, with variance determined by the noise-free image. Thus, also the noise to be injected shall depend on the unknown ground-truth image. To circumvent this issue, we consider the additive injection of noise in variance-stabilized range, where no previous knowledge of the ground-truth signal is necessary. Specifically, we design a special noise-injection operator that prevents the errors on expectation and variance that would otherwise arise when standard variance-stabilizing transformations are used for this task. Thus, the proposed operator is suitable for accurately injecting signal-dependent noise even to images acquired at very low counts.",
keywords = "Anscombe transformation, Noise injection, optimization, Poisson noise, variance stabilization",
author = "Lucas Borges and Marcelo Vieira and Alessandro Foi",
year = "2016",
doi = "10.1109/LSP.2016.2601689",
language = "English",
volume = "23",
pages = "1494--1498",
journal = "IEEE Signal Processing Letters",
issn = "1070-9908",
publisher = "Institute of Electrical and Electronics Engineers",
number = "10",

}

RIS (suitable for import to EndNote) - Lataa

TY - JOUR

T1 - Unbiased Injection of Signal-Dependent Noise in Variance-Stabilized Range

AU - Borges, Lucas

AU - Vieira, Marcelo

AU - Foi, Alessandro

PY - 2016

Y1 - 2016

N2 - The design, optimization, and validation of many image processing or image-based analysis systems often requires testing of the system performance over a dataset of images corrupted by noise at different signal-to-noise ratio regimes. A noise-free ground-truth image may not be available, and different SNRs are simulated by injecting extra noise into an already noisy image. However, noise in real-world systems is typically signal-dependent, with variance determined by the noise-free image. Thus, also the noise to be injected shall depend on the unknown ground-truth image. To circumvent this issue, we consider the additive injection of noise in variance-stabilized range, where no previous knowledge of the ground-truth signal is necessary. Specifically, we design a special noise-injection operator that prevents the errors on expectation and variance that would otherwise arise when standard variance-stabilizing transformations are used for this task. Thus, the proposed operator is suitable for accurately injecting signal-dependent noise even to images acquired at very low counts.

AB - The design, optimization, and validation of many image processing or image-based analysis systems often requires testing of the system performance over a dataset of images corrupted by noise at different signal-to-noise ratio regimes. A noise-free ground-truth image may not be available, and different SNRs are simulated by injecting extra noise into an already noisy image. However, noise in real-world systems is typically signal-dependent, with variance determined by the noise-free image. Thus, also the noise to be injected shall depend on the unknown ground-truth image. To circumvent this issue, we consider the additive injection of noise in variance-stabilized range, where no previous knowledge of the ground-truth signal is necessary. Specifically, we design a special noise-injection operator that prevents the errors on expectation and variance that would otherwise arise when standard variance-stabilizing transformations are used for this task. Thus, the proposed operator is suitable for accurately injecting signal-dependent noise even to images acquired at very low counts.

KW - Anscombe transformation

KW - Noise injection

KW - optimization

KW - Poisson noise

KW - variance stabilization

U2 - 10.1109/LSP.2016.2601689

DO - 10.1109/LSP.2016.2601689

M3 - Article

VL - 23

SP - 1494

EP - 1498

JO - IEEE Signal Processing Letters

JF - IEEE Signal Processing Letters

SN - 1070-9908

IS - 10

ER -