Multi-Noise Removal from Images by Wavelet-based Bayesian Estimator

Xu Huang, A Madoc, Andrew Cheetham

    Research output: A Conference proceeding or a Chapter in BookConference contribution

    5 Citations (Scopus)

    Abstract

    Images are in many cases degraded even before they are encoded. The major noise sources, in terms of distributions, are Gaussian noise and Poisson noise. Noise acquired by images during transmission would be Gaussian in distribution, while images such as emission and transmission tomography images, X-ray films, and photographs taken by satellites are usually contaminated by quantum noise, which is Poisson distributed. Poisson shot noise is a natural generalization of a compound Poisson process when the summands are stochastic processes starting at the points of the underlying Poisson process. Unlike additive Gaussian noise, Poisson noise is signal-dependent and separating signal from noise is a difficult task. In our previous papers we discussed a wavelet-based maximum likelihood for Bayesian estimator that recovers the signal component of wavelet coefficients in original images using an alpha-stable signal prior distribution. In this paper, it is demonstrated that the method can be extended to multi-noise sources comprising both Gaussian and Poisson distributions. Results of varying the parameters of the Bayesian estimators of the model are presented after an investigation of a-stable simulations for a maximum likelihood estimator. As an example, a colour image is processed and presented to illustrate our discussion.
    Original languageEnglish
    Title of host publicationProceedings of the IEEE Sixth International Symposium on Multimedia Software Engineering (ISMSE’04)
    EditorsBob Werner
    Place of PublicationUSA
    PublisherIEEE, Institute of Electrical and Electronics Engineers
    Pages258-264
    Number of pages7
    ISBN (Print)0-7695-2217-3
    DOIs
    Publication statusPublished - 2004
    EventIEEE Sixth International Symposium on Multimedia Software Engg - Florida, United States
    Duration: 13 Dec 200415 Dec 2004

    Conference

    ConferenceIEEE Sixth International Symposium on Multimedia Software Engg
    CountryUnited States
    CityFlorida
    Period13/12/0415/12/04

    Fingerprint

    estimators
    poisson process
    random noise
    stochastic processes
    shot noise
    photographs
    normal density functions
    tomography
    color
    coefficients
    x rays
    simulation

    Cite this

    Huang, X., Madoc, A., & Cheetham, A. (2004). Multi-Noise Removal from Images by Wavelet-based Bayesian Estimator. In B. Werner (Ed.), Proceedings of the IEEE Sixth International Symposium on Multimedia Software Engineering (ISMSE’04) (pp. 258-264). USA: IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/MMSE.2004.53
    Huang, Xu ; Madoc, A ; Cheetham, Andrew. / Multi-Noise Removal from Images by Wavelet-based Bayesian Estimator. Proceedings of the IEEE Sixth International Symposium on Multimedia Software Engineering (ISMSE’04). editor / Bob Werner. USA : IEEE, Institute of Electrical and Electronics Engineers, 2004. pp. 258-264
    @inproceedings{a014dcf1e5894a04beb4b51443e21630,
    title = "Multi-Noise Removal from Images by Wavelet-based Bayesian Estimator",
    abstract = "Images are in many cases degraded even before they are encoded. The major noise sources, in terms of distributions, are Gaussian noise and Poisson noise. Noise acquired by images during transmission would be Gaussian in distribution, while images such as emission and transmission tomography images, X-ray films, and photographs taken by satellites are usually contaminated by quantum noise, which is Poisson distributed. Poisson shot noise is a natural generalization of a compound Poisson process when the summands are stochastic processes starting at the points of the underlying Poisson process. Unlike additive Gaussian noise, Poisson noise is signal-dependent and separating signal from noise is a difficult task. In our previous papers we discussed a wavelet-based maximum likelihood for Bayesian estimator that recovers the signal component of wavelet coefficients in original images using an alpha-stable signal prior distribution. In this paper, it is demonstrated that the method can be extended to multi-noise sources comprising both Gaussian and Poisson distributions. Results of varying the parameters of the Bayesian estimators of the model are presented after an investigation of a-stable simulations for a maximum likelihood estimator. As an example, a colour image is processed and presented to illustrate our discussion.",
    author = "Xu Huang and A Madoc and Andrew Cheetham",
    year = "2004",
    doi = "10.1109/MMSE.2004.53",
    language = "English",
    isbn = "0-7695-2217-3",
    pages = "258--264",
    editor = "Bob Werner",
    booktitle = "Proceedings of the IEEE Sixth International Symposium on Multimedia Software Engineering (ISMSE’04)",
    publisher = "IEEE, Institute of Electrical and Electronics Engineers",
    address = "United States",

    }

    Huang, X, Madoc, A & Cheetham, A 2004, Multi-Noise Removal from Images by Wavelet-based Bayesian Estimator. in B Werner (ed.), Proceedings of the IEEE Sixth International Symposium on Multimedia Software Engineering (ISMSE’04). IEEE, Institute of Electrical and Electronics Engineers, USA, pp. 258-264, IEEE Sixth International Symposium on Multimedia Software Engg, Florida, United States, 13/12/04. https://doi.org/10.1109/MMSE.2004.53

    Multi-Noise Removal from Images by Wavelet-based Bayesian Estimator. / Huang, Xu; Madoc, A; Cheetham, Andrew.

    Proceedings of the IEEE Sixth International Symposium on Multimedia Software Engineering (ISMSE’04). ed. / Bob Werner. USA : IEEE, Institute of Electrical and Electronics Engineers, 2004. p. 258-264.

    Research output: A Conference proceeding or a Chapter in BookConference contribution

    TY - GEN

    T1 - Multi-Noise Removal from Images by Wavelet-based Bayesian Estimator

    AU - Huang, Xu

    AU - Madoc, A

    AU - Cheetham, Andrew

    PY - 2004

    Y1 - 2004

    N2 - Images are in many cases degraded even before they are encoded. The major noise sources, in terms of distributions, are Gaussian noise and Poisson noise. Noise acquired by images during transmission would be Gaussian in distribution, while images such as emission and transmission tomography images, X-ray films, and photographs taken by satellites are usually contaminated by quantum noise, which is Poisson distributed. Poisson shot noise is a natural generalization of a compound Poisson process when the summands are stochastic processes starting at the points of the underlying Poisson process. Unlike additive Gaussian noise, Poisson noise is signal-dependent and separating signal from noise is a difficult task. In our previous papers we discussed a wavelet-based maximum likelihood for Bayesian estimator that recovers the signal component of wavelet coefficients in original images using an alpha-stable signal prior distribution. In this paper, it is demonstrated that the method can be extended to multi-noise sources comprising both Gaussian and Poisson distributions. Results of varying the parameters of the Bayesian estimators of the model are presented after an investigation of a-stable simulations for a maximum likelihood estimator. As an example, a colour image is processed and presented to illustrate our discussion.

    AB - Images are in many cases degraded even before they are encoded. The major noise sources, in terms of distributions, are Gaussian noise and Poisson noise. Noise acquired by images during transmission would be Gaussian in distribution, while images such as emission and transmission tomography images, X-ray films, and photographs taken by satellites are usually contaminated by quantum noise, which is Poisson distributed. Poisson shot noise is a natural generalization of a compound Poisson process when the summands are stochastic processes starting at the points of the underlying Poisson process. Unlike additive Gaussian noise, Poisson noise is signal-dependent and separating signal from noise is a difficult task. In our previous papers we discussed a wavelet-based maximum likelihood for Bayesian estimator that recovers the signal component of wavelet coefficients in original images using an alpha-stable signal prior distribution. In this paper, it is demonstrated that the method can be extended to multi-noise sources comprising both Gaussian and Poisson distributions. Results of varying the parameters of the Bayesian estimators of the model are presented after an investigation of a-stable simulations for a maximum likelihood estimator. As an example, a colour image is processed and presented to illustrate our discussion.

    U2 - 10.1109/MMSE.2004.53

    DO - 10.1109/MMSE.2004.53

    M3 - Conference contribution

    SN - 0-7695-2217-3

    SP - 258

    EP - 264

    BT - Proceedings of the IEEE Sixth International Symposium on Multimedia Software Engineering (ISMSE’04)

    A2 - Werner, Bob

    PB - IEEE, Institute of Electrical and Electronics Engineers

    CY - USA

    ER -

    Huang X, Madoc A, Cheetham A. Multi-Noise Removal from Images by Wavelet-based Bayesian Estimator. In Werner B, editor, Proceedings of the IEEE Sixth International Symposium on Multimedia Software Engineering (ISMSE’04). USA: IEEE, Institute of Electrical and Electronics Engineers. 2004. p. 258-264 https://doi.org/10.1109/MMSE.2004.53