Low rank prior in single patches for non-pointwise impulse noise removal

Ruixuan Wang (Lead / Corresponding author), Markus Pakleppa, Emanuele Trucco

    Research output: Contribution to journalArticlepeer-review

    15 Citations (Scopus)
    486 Downloads (Pure)

    Abstract

    This paper introduces a low rank prior in small oriented noise-free image patches: Considering an oriented patch as a matrix, a low-rank matrix approximation is enough to preserve the texture details in the optimally oriented patch. Based on this prior, we propose a single-patch method within a generalized joint low-rank and sparse matrix recovery framework to simultaneously detect and remove non-pointwise random-valued impulse noise (e.g., very small blobs). A weighting matrix is incorporated in the framework to encode an initial estimate of the spatial noise distribution. An accelerated proximal gradient method is adapted to estimate the optimal noise-free image patches. Experiments show the effectiveness of our framework in detecting and removing non-pointwise random-valued impulse noise.
    Original languageEnglish
    Pages (from-to)1485-1496
    Number of pages12
    JournalIEEE Transactions on Image Processing
    Volume24
    Issue number5
    Early online date4 Feb 2015
    DOIs
    Publication statusPublished - 2015

    Keywords

    • Approximation methods
    • Equations
    • Image edge detection
    • Sparse matrices
    • Noise reduction
    • Noise
    • Noise measurement

    Fingerprint Dive into the research topics of 'Low rank prior in single patches for non-pointwise impulse noise removal'. Together they form a unique fingerprint.

    Cite this