A Maximum a Posteriori Estimation Framework for Robust High Dynamic Range Video Synthesis

Yuelong Li, Chul Lee, Vishal Monga

    Research output: Contribution to journalArticle

    5 Citations (Scopus)

    Abstract

    High dynamic range (HDR) image synthesis from multiple low dynamic range exposures continues to be actively researched. The extension to HDR video synthesis is a topic of significant current interest due to potential cost benefits. For HDR video, a stiff practical challenge presents itself in the form of accurate correspondence estimation of objects between video frames. In particular, loss of data resulting from poor exposures and varying intensity makes conventional optical flow methods highly inaccurate. We avoid exact correspondence estimation by proposing a statistical approach via maximum a posterior estimation, and under appropriate statistical assumptions and choice of priors and models, we reduce it to an optimization problem of solving for the foreground and background of the target frame. We obtain the background through rank minimization and estimate the foreground via a novel multiscale adaptive kernel regression technique, which implicitly captures local structure and temporal motion by solving an unconstrained optimization problem. Extensive experimental results on both real and synthetic data sets demonstrate that our algorithm is more capable of delivering high-quality HDR videos than current state-of-the-art methods, under both subjective and objective assessments. Furthermore, a thorough complexity analysis reveals that our algorithm achieves better complexity-performance tradeoff than conventional methods.

    Original languageEnglish (US)
    Article number7792731
    Pages (from-to)1143-1157
    Number of pages15
    JournalIEEE Transactions on Image Processing
    Volume26
    Issue number3
    DOIs
    StatePublished - Mar 1 2017

    Fingerprint

    Optical flows
    Costs

    All Science Journal Classification (ASJC) codes

    • Software
    • Computer Graphics and Computer-Aided Design

    Cite this

    @article{1907a6e43e96404487a081b5bd53b45f,
    title = "A Maximum a Posteriori Estimation Framework for Robust High Dynamic Range Video Synthesis",
    abstract = "High dynamic range (HDR) image synthesis from multiple low dynamic range exposures continues to be actively researched. The extension to HDR video synthesis is a topic of significant current interest due to potential cost benefits. For HDR video, a stiff practical challenge presents itself in the form of accurate correspondence estimation of objects between video frames. In particular, loss of data resulting from poor exposures and varying intensity makes conventional optical flow methods highly inaccurate. We avoid exact correspondence estimation by proposing a statistical approach via maximum a posterior estimation, and under appropriate statistical assumptions and choice of priors and models, we reduce it to an optimization problem of solving for the foreground and background of the target frame. We obtain the background through rank minimization and estimate the foreground via a novel multiscale adaptive kernel regression technique, which implicitly captures local structure and temporal motion by solving an unconstrained optimization problem. Extensive experimental results on both real and synthetic data sets demonstrate that our algorithm is more capable of delivering high-quality HDR videos than current state-of-the-art methods, under both subjective and objective assessments. Furthermore, a thorough complexity analysis reveals that our algorithm achieves better complexity-performance tradeoff than conventional methods.",
    author = "Yuelong Li and Chul Lee and Vishal Monga",
    year = "2017",
    month = "3",
    day = "1",
    doi = "10.1109/TIP.2016.2642790",
    language = "English (US)",
    volume = "26",
    pages = "1143--1157",
    journal = "IEEE Transactions on Image Processing",
    issn = "1057-7149",
    publisher = "Institute of Electrical and Electronics Engineers Inc.",
    number = "3",

    }

    A Maximum a Posteriori Estimation Framework for Robust High Dynamic Range Video Synthesis. / Li, Yuelong; Lee, Chul; Monga, Vishal.

    In: IEEE Transactions on Image Processing, Vol. 26, No. 3, 7792731, 01.03.2017, p. 1143-1157.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - A Maximum a Posteriori Estimation Framework for Robust High Dynamic Range Video Synthesis

    AU - Li, Yuelong

    AU - Lee, Chul

    AU - Monga, Vishal

    PY - 2017/3/1

    Y1 - 2017/3/1

    N2 - High dynamic range (HDR) image synthesis from multiple low dynamic range exposures continues to be actively researched. The extension to HDR video synthesis is a topic of significant current interest due to potential cost benefits. For HDR video, a stiff practical challenge presents itself in the form of accurate correspondence estimation of objects between video frames. In particular, loss of data resulting from poor exposures and varying intensity makes conventional optical flow methods highly inaccurate. We avoid exact correspondence estimation by proposing a statistical approach via maximum a posterior estimation, and under appropriate statistical assumptions and choice of priors and models, we reduce it to an optimization problem of solving for the foreground and background of the target frame. We obtain the background through rank minimization and estimate the foreground via a novel multiscale adaptive kernel regression technique, which implicitly captures local structure and temporal motion by solving an unconstrained optimization problem. Extensive experimental results on both real and synthetic data sets demonstrate that our algorithm is more capable of delivering high-quality HDR videos than current state-of-the-art methods, under both subjective and objective assessments. Furthermore, a thorough complexity analysis reveals that our algorithm achieves better complexity-performance tradeoff than conventional methods.

    AB - High dynamic range (HDR) image synthesis from multiple low dynamic range exposures continues to be actively researched. The extension to HDR video synthesis is a topic of significant current interest due to potential cost benefits. For HDR video, a stiff practical challenge presents itself in the form of accurate correspondence estimation of objects between video frames. In particular, loss of data resulting from poor exposures and varying intensity makes conventional optical flow methods highly inaccurate. We avoid exact correspondence estimation by proposing a statistical approach via maximum a posterior estimation, and under appropriate statistical assumptions and choice of priors and models, we reduce it to an optimization problem of solving for the foreground and background of the target frame. We obtain the background through rank minimization and estimate the foreground via a novel multiscale adaptive kernel regression technique, which implicitly captures local structure and temporal motion by solving an unconstrained optimization problem. Extensive experimental results on both real and synthetic data sets demonstrate that our algorithm is more capable of delivering high-quality HDR videos than current state-of-the-art methods, under both subjective and objective assessments. Furthermore, a thorough complexity analysis reveals that our algorithm achieves better complexity-performance tradeoff than conventional methods.

    UR - http://www.scopus.com/inward/record.url?scp=85015175820&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85015175820&partnerID=8YFLogxK

    U2 - 10.1109/TIP.2016.2642790

    DO - 10.1109/TIP.2016.2642790

    M3 - Article

    VL - 26

    SP - 1143

    EP - 1157

    JO - IEEE Transactions on Image Processing

    JF - IEEE Transactions on Image Processing

    SN - 1057-7149

    IS - 3

    M1 - 7792731

    ER -