Method for confirming initial quantization step in code rate control

Abstract

The invention relates to a method for confirming the initial quantization step in code rate control, which comprises the following steps: calculating the quantization step by utilizing the image feature of a first image; mapping the quantization step into a quantization parameter; and coding the first image by taking the quantization parameter as the initial quantization step. The method is suitable for images with different complexity, thereby enhancing the coding quality. The method can be applied to various video coding standards.

Description

technical field [0001] The invention relates to video coding, in particular to a method for determining initial QP in code rate control. Background technique [0002] In order to save transmission bandwidth and storage space, support applications such as digital TV broadcasting, remote monitoring, digital video on demand, wireless multimedia communication, etc., video coding has become one of the hotspots in research and industrial applications at home and abroad. So far, a series of video coding standards have been formulated internationally, such as MPEG-1 / 2 / 4, H.261 / H.263, H.264 / AVC, and the domestic AVS video coding standard is currently being developed and improved. Among them, compared with the previous standards, H.264 / AVC and AVS have similar coding frameworks, but their compression efficiency is at least double that of any previous standard, and are suitable for application in more scenarios. figure 1 Encoding framework for H.264 / AVC and AVS. [0003] The main cod...

AI Technical Summary

Problems solved by technology

[0021] However, the initial QP used in the first step of the current rate control does not consider the characteristics of the image, and only selects the initial QP from bpp, which leads to the following three problems:

If the rate control selects a smaller initial QP, it will cause the length of the encoded code stream to be too large, exceeding the upper limit of the buffer, and the client buffer overflow occurs; conversely, if the rate control selects a larger initial QP, then The length of the encoded code stream will be too small, exceeding the lower limit of the buffer, and the client buffer underflow will occur

[0029] Figure 5 It shows the status of the client buffer when the AVS reference software encodes the Stefan sequence. It can be seen that because the initial QP is too small, the client buffer overflows at the beginning of the encoding (above the red dotted line indicates overflow)

[0030] In the end, it leads to large fluctuations in PSNR and a decrease in overall quality:

Images