Rate distortion optimization quantization method and device, storage medium and terminal

Abstract

The invention discloses a rate distortion optimization quantization method. The method comprises the following steps: generating a pre-quantization coefficient set according to a to-be-quantized transform block; obtaining an nth pre-quantization coefficient from the pre-quantization coefficient set, and generating three alternative coefficients of the nth pre-quantization coefficient; calculating respective indexes of the alternative coefficients according to the (n-1)th pre-quantization coefficient of the nth pre-quantization coefficient, and searching respective code rates of the alternative coefficients according to the respective indexes; calculating the respective rate distortion cost of the alternative coefficients according to the respective code rates of the alternative coefficients, and determining the alternative coefficient with the minimum rate distortion cost as the nth optimal quantization coefficient; continuing to execute the step of obtaining the nth pre-quantization coefficient from the pre-quantization coefficient set until the traversal of each pre-quantization coefficient in the pre-quantization coefficient set is finished, and generating an optimal quantization coefficient set; and determining a quantization result of the to-be-quantized transform block based on the optimal quantization coefficient set, wherein the formula for obtaining the nth pre-quantization coefficient is n=k, and k is the number of obtaining times. Therefore, the coding efficiency of the video image can be improved.

Description

technical field [0001] The present invention relates to the technical field of digital signal processing, in particular to a rate-distortion optimized quantization method, device, storage medium and terminal. Background technique [0002] In the current image and video coding technology, compression is mainly achieved by removing spatial redundancy, temporal redundancy (in the video field) and coding redundancy. The encoder effectively combines multiple algorithms to achieve high compression performance. Related techniques include predictive coding, transform coding, quantization, and entropy coding. The quantization is usually performed in a manner of rate-distortion optimized quantization. [0003] Rate Distortion Optimization Quantization (RateDistortionOptimizationQuantization, RDOQ) is a very important technology in the AVS3 video coding standard, it can effectively improve the performance of the encoder, but the RDOQ module needs to calculate a large number of rate-di...

AI Technical Summary

Problems solved by technology

However, the calculation of the table index also has dependencies, such as calculating figure 1 As shown, when calculating the table index of several candidate coefficients corresponding to the current coefficient, it is calculated based on the optimal quantization coefficient determined by the previous coefficient of the current coefficient. That is to say, when calculating the index of the current coefficient, it must Waiting for the optimal quantization coefficient determined by the previous coefficient to be calculated (that is, the serial calculation method), which leads to an increase in encoding time and reduces the encoding efficiency of video image data

Images