Loop filtering method in image coding processing

Abstract

A loop filtering method for use in image encoding process includes: determination step of the block effect value, for determining the block effect value of the coding block according to the amount of the edge area similar point of coding block to be filtered; judgement step of the real edge of image, for judging whether the edge of the coding block is the real edge of image according to the pixel difference value of the two edge sides of the coding blocks; filtering step, for filtering the coding block in the filtering mode corresponding the block effect value of the coding block when the edge of the coding block is not the real edge of image. By said steps, the S / N ratio can be increased by about 0.08db, and the code rate can be decreased by about 1.5%.

Description

technical field [0001] The present invention relates to the technical field of image coding processing, in particular to a loop filtering method in image coding processing. Background technique [0002] With the development of multimedia technology and the rapid promotion of multimedia applications, the H.264 video image compression standard formulated by JVT, a standard formulation working group jointly formed by MPEG and ITU, is characterized by its advanced technical characteristics and better compression It has been widely used in conference TV, videophone, streaming media, HD DVD, video surveillance, digital TV, 3G and other fields. [0003] H.264 compression processing framework such as figure 1 As shown, the basic processing unit is 16×16 macroblock, which adopts multi-frame reference, intra-frame prediction, multi-macroblock mode, 4×4 integer transformation and quantization, loop filtering, 1 / 4 pixel motion prediction, CAVLC ( Context-based variable length coding a...

AI Technical Summary

Problems solved by technology

[0017] It can be seen that the above-mentioned existing loop filtering process ignores the fact that the block effect is easier to be perceived by the human eye in the flat area, so it does not perform high-intensity smoothing filtering for the flat area

Moreover, in the above method, the filtering operator has only 3 levels (0, 1, 2). Since the filtering operator is too single, it is difficult to achieve a good filtering effect, which makes the corresponding method suitable for limited occasions.

Images