Loading patent…
AssigneeThomson Licensing
InventorsPhilippe Bordes, Yannick Olivier, David Touze, Franck Hiron
Estimating the term…
In order to take into account the local illumination variations in a scene, the invention proposes a method of compensating illumination variations in which distributions are determined for each block of an image (or view) and, in order to avoid a loss of the video coding efficiency, a distribution-based mapping function is automatically computed by a remote apparatus, such a decoder, from only information available at that apparatus when decoding a current block of the sequence of images. The present invention relates also to a method and device of encoding and/or decoding a sequence of images which comprises means configured to implement the method of compensating illumination variations according to the invention.
A method of compensating illumination variations in a sequence of images, wherein, for each current block of a current image from the sequence of images, the method comprising:
An apparatus for compensating illumination variations in a sequence of images which comprises a memory associated with a processor, said memory and processor configured to determine:
A device for encoding a sequence of images, comprising an apparatus comprising a memory associated with a processor, said memory and processor configured to send a signal and obtain, from said sent signal, an information about a mapping between pixel values of a block of a current image from a sequence of images and pixel values of a block of a reference image from the sequence of images obtained by matching two distributions, one computed over a causal region surrounding said block of the current image and the other one over a causal region surrounding said block of the reference image.
A device for decoding a sequence of images, comprising a memory associated with a processor, said memory and processor configured to receive a signal and obtain, from said received signal, an information about a mapping between pixel values of a block of a current image from a sequence of images and pixel values of a block of a reference image from the sequence of images obtained by matching two distributions, one computed over a causal region surrounding said block of the current image and the other one over a causal region surrounding said block of the reference image.
Dependencies are inferred from the claim text.