A coding method to improve the subjective quality of hdr video

Abstract

The invention discloses an encoding method for improving the subjective quality of an HDR video. The method specifically comprises the following steps that firstly, quantization parameters are obtained by using video source data through conventional code rate control; secondly, a texture evaluation value and a brightness evaluation value of a current encoding unit of the video source data are obtained; thirdly, the texture evaluation value and the brightness evaluation value are fused into a texture brightness measurement factor; fourthly, according to the quantization parameters in the first step and the texture brightness measurement factor in the third step, quantization parameters of the current encoding unit is obtained, and the video source data is quantified. The method has the advantages that the subjective quality of a basic layer area with a large brightness range is not reduced, and detail layers in bright and dark areas with a small brightness range are kept more complete, so that the subjective quality of the whole HDR video is effectively improved.

Description

technical field [0001] The present invention relates to the technical field of video coding, in particular to a coding method for improving the subjective quality of HDR video. Background technique [0002] With the rapid development of multimedia information technology, HDR (High Dynamic Range, high dynamic range) technology has become a hot spot in the video industry, and hardware devices such as HDR cameras, HDR set-top boxes, and HDR TVs have begun to appear in the mass consumer market. Compared with traditional SDR (Standard Dynamic Range, standard dynamic range) video, HDR video can show wider brightness and more colors, and it will not overexpose when it is bright, and it will not be underexposed when it is dark. The outline of the object can still be clearly distinguished in the dark and dark places, so it can better show the real content of the video. [0003] HDR video mainly goes through the following four steps from shooting to presenting to the audience: [00...

AI Technical Summary

Problems solved by technology

[0012] The general HDR encoding on the market adopts the original HEVC (High Efficiency Video Coding) scheme for SDR video encoding. This scheme does not take into account the characteristics of HDR video. It determines the size of quantization parameters according to the complexity of each area of ​​the image. Less consideration of the brightness factor will easily lead to brighter and darker areas during encoding, which will cause greater distortion due to excessive quantization, and a lot of information in the detail layer will be lost.

Images