A method, device and computer-readable storage medium for removing block effect

Abstract

The present application discloses a method, device, and computer-readable storage medium for removing block effects. The method includes: performing edge enhancement processing and non-edge smoothing processing on the current image, generating the current image to be encoded and dividing it into multiple encoding blocks; Select a prediction mode from multiple prediction modes as the prediction mode to be selected; judge whether the prediction mode to be selected is a prediction mode that generates block effects; if the prediction mode to be selected is a prediction mode that generates block effects, then exist When there is an edge, return to select a prediction mode from multiple prediction modes as the candidate prediction mode until the candidate prediction mode is used to predict the coding block without blockiness or there is no edge in the coding block; the candidate prediction mode It is recorded as the current prediction mode, and the current image to be encoded is processed by using the current prediction mode to obtain the coded stream. Through the above method, the present application can remove the blocking effect and improve the coding quality.

Description

Technical field [0001] The present application relates to the field of image processing technology, particularly relates to a method of removing blocking artifacts, apparatus, and computer-readable storage medium. Background technique [0002] Current mainstream video image / video codecs are based coding mode and a quantization block, the block-based encoding of the entire image may be divided so that more accurate prediction, coding and facilitates parallel speedup. Since all of the image blocks are independently coded, quantization process is not dependent on the surrounding blocks, resulting in the quantized coefficient block and the current block o uncorrelated, and further such that the two adjacent pixels at the boundary appears there is discontinuous to produce diamond effect, this phenomenon is particularly evident especially at low bit rates, while the human eye is very sensitive, leading to subjective visual quality deteriorates. Although Some codecs using standard tec...

AI Technical Summary

Problems solved by technology

Since all blocks in the image are encoded independently, the quantization process does not depend on the surrounding blocks, resulting in the irrelevance of the quantization coefficients of adjacent blocks and the current block, so that the pixels at the boundary of two adjacent blocks will be discontinuous. , resulting in block effects, especially at low bit rates, and the human eye is very sensitive to this, resulting in poor subjective visual quality

Although some codec standard technologies currently use some post-processing methods (such as: block filtering, sample adaptive compensation or adaptive filtering, etc.) to process the reconstructed image to make it closer to the original image, the block effect is still serious, especially It is for the case of low code rate

Images