Video macroblock context information access method and system

Abstract

The invention provides a video macroblock context information access method and system. The method includes the steps that when a coder decoder codes or decodes a current macroblock, the coder decoder reads context information of an adjacent macroblock stored in a macroblock context memorizer and context information of a sub-macroblock in the current macroblock stored in a sub-macroblock context memorizer; the coder decoder codes or decodes the current macroblock according to the context information of the adjacent macroblock and the context information of the sub-macroblock; during coding and decoding of the current macroblock, the context information of the sub-macroblcok is updated and stored in the sub-macroblock context memorizer; after coding and decoding of the current macroblock, the context information of the current macroblock is extracted and stored in the macroblock context memorizer.

Description

technical field [0001] The present invention relates to a method and system for accessing context information of video macroblocks, in particular to a method and system for accessing context information of video macroblocks designed based on entropy codec-oriented context structure design of video macroblocks. Background technique [0002] In existing video codec standards, codecs based on entropy codecs generally use predictive codecs extensively, that is, use context information of adjacent macroblocks and context information of sub-macroblocks inside macroblocks as codec predictive values, and continuously update these contextual information as the encoding and decoding progresses. The use of these context information improves codec efficiency and also increases the complexity of codec design. For example, for some mainstream video standards such as H.264, the minimum unit corresponding to context information reaches the level of 4x4 sub-macroblocks. That is to say, the ...

AI Technical Summary

Problems solved by technology

This method saves all the context information inside the encoding and decoding hardware (in-chip), and completely accesses all the required context information by the hardware, which will greatly improve the execution efficiency, but the size and complexity of the hardware due to storage space and addressing increase, which makes the hardware design cost larger and the cost is high

Images