Implementation of Discrete Cosine Transform Based on High Efficiency Video Coding Standard

Abstract

The invention provides a discrete cosine transform implementation method based on an efficient video encoding standard. The specific implementation steps are as follows: (1) obtaining image residual pixels; (2) reading a residual pixel matrix; (3) reading any line of pixel values; (4) performing 32-point butterfly transform; (5) performing16-point butterfly transform; (6) performing 8-point butterfly transform; (7) performing 4-point discrete cosine transform; (8) adjusting the sequence of pixel values; (9) determining whether the last line is processed; (10) performing two-dimensional discrete cosine transform; (11) judging whether the last matrix is processed; and (12) outputting code streams. By adoption of the discrete cosine transform implementation method provided by the invention, the problem of relatively high hardware cost of the discrete cosine transform implementation method in the prior art can be effectively overcome, the hardware cost in an encoding process is reduced, and images can be well compressed.

Description

technical field [0001] The invention belongs to the technical field of image coding, and further relates to a discrete cosine transform implementation method based on high-efficiency video coding standards in the technical field of image compression. The invention can be used for image and video compression encoding in the fields of network image processing, satellite image transmission and the like. Background technique [0002] As a new generation of video coding standard, the high-efficiency video coding standard mainly improves the compression of high-definition and ultra-high-definition resolution videos. Compared with the previous H.264 / AVC standard, under the same compression effect, the code stream size only half of the former. The high-efficiency video coding standard is divided into two parts: intra-frame coding and inter-frame coding. For intra-frame coding, it uses discrete cosine transform to realize the compression processing of video pixels, so that the scatt...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to address above-mentioned deficiencies in the prior art, proposes a kind of discrete cosine transform realization method based on high-efficiency video coding standard, adopts multiplexing structure, addition displacement to replace multiplier and use random access memory SRAM to store intermediate data method to reduce hardware overhead, solve the problem of high power consumption of the encoder hardware, and increase the operating frequency of the encoder

Images