Video coding using transforms bigger than 4x4 and 8x8

Video coding using transforms bigger than 4x4 and 8x8

  • CN 102,204,251 A
  • Filed: 09/30/2009
  • Published: 09/28/2011
  • Est. Priority Date: 10/03/2008
  • Status: Active Grant
First Claim
Patent Images

1. the method for a coding video frequency data, described method comprises:

  • Based on predictive mode with spatial prediction or motion compensation applications the original video block in frame of video to produce predicted video block;

    Described original video block in described frame of video deducts described predicted video block to form residual block;

    Selection has the conversion of first transform size to be applied to described residual block;

    Produce the header data of the described selected transform of indication, described header data comprises first syntactic element of first value with at least one transform size of indication and indicates second syntactic element of the prediction block sizes of described predicted video block, described first transform size of the common indication of wherein said first syntactic element and described second syntactic element;

    Described selected transform is applied to described residual block to produce remaining conversion coefficient;

    AndProduce vision signal based on described header data and described remaining conversion coefficient.

View all claims

    Thank you for your feedback