Loop filter method and loop filter system

Abstract

The invention provides a loop filter method and a loop filter system. The method comprises the following steps: filtering macro blocks one by one; for each sub block inside the micro blocks, firstly filtering the horizontal edge of the sub block, and then filtering the vertical edge of the sub block; or firstly filtering the vertical edge of the sub block, and then filtering the horizontal edge of the sub block; filtering the horizontal edge of each sub block of each row of sub blocks in order from top to bottom and from left to right; and filtering the vertical edge of each sub block of eachrow of sub blocks in order from left to right and from top to bottom. The invention also provides the loop filter system. According to the loop filter method of the invention, the storage number of macro data is reduced, and the memory space is saved; and the method reduces the reading times of the sub blocks and the storage time of the data, and improves the reading rate; moreover, the complexity of hardware design is reduced by adopting a regular filter order, and the hardware overhead is low.

Description

technical field [0001] The invention relates to the technical field of digital video decoding, in particular to a loop filtering method and a loop filtering system. Background technique [0002] Hardware decoders usually decode macroblocks in a pipelined manner, and the decoding order is from left to right and from top to bottom. Take a 64×64 pixel image as an example, which includes 4×4 macroblocks, and each macroblock contains 4×4 sub-blocks. According to the decoding order of RV9, its hardware filtering order is as follows figure 1 shown. The number on the left in the figure is the number of the macroblock, U refers to the upper 3 rows of subblocks of the macroblock, L refers to the lower row of subblocks of the macroblock, and the numbers in brackets are the hardware filtering order. [0003] The upper-level pipeline decodes a macroblock and sends it to the filter for filtering. The filter must store the data of this macroblock in its own memory. For example, after th...

AI Technical Summary

Problems solved by technology

[0007] From the above analysis, it can be seen that according to the filtering sequence of RV9, in order to implement filtering by hardware, more sub-block data needs to be saved, and because the data is read more times, the calculation time is relatively long, and at the same time, this irregular The calculation order of will increase the complexity of hardware design

Images