Cepstral analysis based blind deconvolution for motion blur

Haruka Asai, Yuji Oyamada, Julien Pilet, Hideo Saito

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    11 Citations (Scopus)

    Abstract

    Camera shake during exposure blurs the captured image. Despite several decades of studies, image deconvolution to restore a blurred image still remains an issue, particularly in blind deconvolution cases in which the actual shape of the blur is unknown. Approaches based on cepstral analysis succeeded in restoring images degraded by a uniform blur caused by a camera moving straight in a single direction. In this paper, we propose to estimate, from a single blurred image, the point spread function (PSF) caused by a normal camera undergoing a 2D curved motion, and to restore the image. To extend the traditional cepstral analysis, we derive assumptions about the PSF effects in the cepstrum domain. In a first phase, we estimate several PSF candidates from the cepstrum of a blurred image and restore the image with a fast deconvolution algorithm. In a second phase, we select the best PSF candidate by evaluating the restored images. Finally, a slower but more accurate deconvolution algorithm recovers the latent image with the chosen PSF. We validate the proposed method with synthetic and real experiments.

    Original languageEnglish
    Title of host publicationProceedings - International Conference on Image Processing, ICIP
    Pages1153-1156
    Number of pages4
    DOIs
    Publication statusPublished - 2010
    Event2010 17th IEEE International Conference on Image Processing, ICIP 2010 - Hong Kong
    Duration: 2010 Sep 262010 Sep 29

    Other

    Other2010 17th IEEE International Conference on Image Processing, ICIP 2010
    CityHong Kong
    Period10/9/2610/9/29

    Fingerprint

    Optical transfer function
    Deconvolution
    Cameras
    Experiments

    Keywords

    • Blind deconvolution
    • Cepstral analysis
    • Image restoration
    • Point spread function

    ASJC Scopus subject areas

    • Software
    • Computer Vision and Pattern Recognition
    • Signal Processing

    Cite this

    Asai, H., Oyamada, Y., Pilet, J., & Saito, H. (2010). Cepstral analysis based blind deconvolution for motion blur. In Proceedings - International Conference on Image Processing, ICIP (pp. 1153-1156). [5651299] https://doi.org/10.1109/ICIP.2010.5651299

    Cepstral analysis based blind deconvolution for motion blur. / Asai, Haruka; Oyamada, Yuji; Pilet, Julien; Saito, Hideo.

    Proceedings - International Conference on Image Processing, ICIP. 2010. p. 1153-1156 5651299.

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Asai, H, Oyamada, Y, Pilet, J & Saito, H 2010, Cepstral analysis based blind deconvolution for motion blur. in Proceedings - International Conference on Image Processing, ICIP., 5651299, pp. 1153-1156, 2010 17th IEEE International Conference on Image Processing, ICIP 2010, Hong Kong, 10/9/26. https://doi.org/10.1109/ICIP.2010.5651299
    Asai H, Oyamada Y, Pilet J, Saito H. Cepstral analysis based blind deconvolution for motion blur. In Proceedings - International Conference on Image Processing, ICIP. 2010. p. 1153-1156. 5651299 https://doi.org/10.1109/ICIP.2010.5651299
    Asai, Haruka ; Oyamada, Yuji ; Pilet, Julien ; Saito, Hideo. / Cepstral analysis based blind deconvolution for motion blur. Proceedings - International Conference on Image Processing, ICIP. 2010. pp. 1153-1156
    @inproceedings{b7a91cdd7a9e4741ae48d0029aef68df,
    title = "Cepstral analysis based blind deconvolution for motion blur",
    abstract = "Camera shake during exposure blurs the captured image. Despite several decades of studies, image deconvolution to restore a blurred image still remains an issue, particularly in blind deconvolution cases in which the actual shape of the blur is unknown. Approaches based on cepstral analysis succeeded in restoring images degraded by a uniform blur caused by a camera moving straight in a single direction. In this paper, we propose to estimate, from a single blurred image, the point spread function (PSF) caused by a normal camera undergoing a 2D curved motion, and to restore the image. To extend the traditional cepstral analysis, we derive assumptions about the PSF effects in the cepstrum domain. In a first phase, we estimate several PSF candidates from the cepstrum of a blurred image and restore the image with a fast deconvolution algorithm. In a second phase, we select the best PSF candidate by evaluating the restored images. Finally, a slower but more accurate deconvolution algorithm recovers the latent image with the chosen PSF. We validate the proposed method with synthetic and real experiments.",
    keywords = "Blind deconvolution, Cepstral analysis, Image restoration, Point spread function",
    author = "Haruka Asai and Yuji Oyamada and Julien Pilet and Hideo Saito",
    year = "2010",
    doi = "10.1109/ICIP.2010.5651299",
    language = "English",
    isbn = "9781424479948",
    pages = "1153--1156",
    booktitle = "Proceedings - International Conference on Image Processing, ICIP",

    }

    TY - GEN

    T1 - Cepstral analysis based blind deconvolution for motion blur

    AU - Asai, Haruka

    AU - Oyamada, Yuji

    AU - Pilet, Julien

    AU - Saito, Hideo

    PY - 2010

    Y1 - 2010

    N2 - Camera shake during exposure blurs the captured image. Despite several decades of studies, image deconvolution to restore a blurred image still remains an issue, particularly in blind deconvolution cases in which the actual shape of the blur is unknown. Approaches based on cepstral analysis succeeded in restoring images degraded by a uniform blur caused by a camera moving straight in a single direction. In this paper, we propose to estimate, from a single blurred image, the point spread function (PSF) caused by a normal camera undergoing a 2D curved motion, and to restore the image. To extend the traditional cepstral analysis, we derive assumptions about the PSF effects in the cepstrum domain. In a first phase, we estimate several PSF candidates from the cepstrum of a blurred image and restore the image with a fast deconvolution algorithm. In a second phase, we select the best PSF candidate by evaluating the restored images. Finally, a slower but more accurate deconvolution algorithm recovers the latent image with the chosen PSF. We validate the proposed method with synthetic and real experiments.

    AB - Camera shake during exposure blurs the captured image. Despite several decades of studies, image deconvolution to restore a blurred image still remains an issue, particularly in blind deconvolution cases in which the actual shape of the blur is unknown. Approaches based on cepstral analysis succeeded in restoring images degraded by a uniform blur caused by a camera moving straight in a single direction. In this paper, we propose to estimate, from a single blurred image, the point spread function (PSF) caused by a normal camera undergoing a 2D curved motion, and to restore the image. To extend the traditional cepstral analysis, we derive assumptions about the PSF effects in the cepstrum domain. In a first phase, we estimate several PSF candidates from the cepstrum of a blurred image and restore the image with a fast deconvolution algorithm. In a second phase, we select the best PSF candidate by evaluating the restored images. Finally, a slower but more accurate deconvolution algorithm recovers the latent image with the chosen PSF. We validate the proposed method with synthetic and real experiments.

    KW - Blind deconvolution

    KW - Cepstral analysis

    KW - Image restoration

    KW - Point spread function

    UR - http://www.scopus.com/inward/record.url?scp=78651069611&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=78651069611&partnerID=8YFLogxK

    U2 - 10.1109/ICIP.2010.5651299

    DO - 10.1109/ICIP.2010.5651299

    M3 - Conference contribution

    SN - 9781424479948

    SP - 1153

    EP - 1156

    BT - Proceedings - International Conference on Image Processing, ICIP

    ER -