The coded block parameters used to code blocks of image samples into structures called macroblocks are compressed more efficiently by exploiting the correlation between chrominance and luminance blocks in each macroblock. In particular, the coded block pattern for chrominance and luminance are combined into a single parameter for the macroblock and jointly coded with a single variable length code. To further enhance coding efficiency, the spatial coherence of coded block patterns can be exploited by using spatial prediction to compute predicted values for coded block pattern parameters.

 
Web www.patentalert.com

< Creating and managing structured data in an electronic spreadsheet

< Pose-invariant face recognition system and process

> Method and apparatus for speech synthesis without prosody modification

> Controlling access to objects with rules for a work management environment

~ 00298