Inter-frame fast video transcoding method applied to h.264 to hevc

Abstract

The invention provides an interframe rapid video transcoding method applied to from H. 264 to HEVC. According to H. 264 code stream information and HEVC interframe coding information, for the interframe coding part in an H. 264 to the HEVC transcoder, a fast mode decision and motion estimation method is adopted. With regard to a CU with the depth of 0, 1 layers, combined with H. 264 code stream decoded information and HEVC coded information, a 2N * 2N in-advance determination threshold and an asymmetric motion partition (AMP) determination threshold are obtained, and then the search condition of its PU is determined. For the motion estimation part, the motion vector search range and the motion vector search precision are determined through the H. 264 decoding information. The motion vector search precision is divided into three parts of the integer pixel precision, the sub-pixel precision and the zero motion vector. The method can greatly reduce the computation complexity of interframe coding in the transcoding process from H. 264 to HEVC.

Description

technical field [0001] The present invention relates to transcoding of video data, in particular to a statistical analysis-based transcoding method from H.264 to HEVC, which is used to transcode video data of H.264 / AVC standard into video data of HEVC standard. Background technique [0002] In 2013, two major international organizations, MPEG and ITU, officially released the HEVC next-generation digital video codec standard to adapt to the development of high-quality video applications. HEVC uses advanced technology to improve the relationship between code stream, coding quality, delay and algorithm complexity, effectively improving video coding efficiency. HEVC is still a hybrid coding framework that follows the prediction and transformation of previous generations of coding standards. Unlike H.264, HEVC enlarges the size of the coding block. In H.264, the encoding block size is from 4×4 to 16×16, while HEVC expands the block size from 4×4 to 128×128 to deal with high-res...

AI Technical Summary

Problems solved by technology

[0007] Document 2 (see Peixoto, E., Izquierdo, E., "A complexity-scalable transcoder from H.264 / AVC to the new HEVC codec," Image Processing (ICIP), 2012 19th IEEEInternational Conference on, pp.737-740, Sept.30 2012-Oct.3 2012.) proposed a motion vector-based transcoding algorithm, through the processing of the motion vector, to obtain the threshold of coding unit division prediction, and to divide or terminate in advance by limiting the threshold , due to the relatively monotonous motion information used, the algorithm brings a huge BD-rate loss while accelerating

Images