H.264 to HEVC Transcoding via Parallel GOP Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high computational complexity of the High Efficiency Video Coding (HEVC) standard makes it difficult to implement on general-purpose processors, limiting its adoption in cloud-based multimedia encoding/transcoding systems and real-time systems, and there is a need for rapid transcoding of existing H.264/AVC bitstreams into HEVC bitstreams due to the large amount of legacy content already encoded with the H.264/AVC standard.
Innovation Solution
A method and system for transcoding H.264/AVC bitstreams into HEVC bitstreams by parsing the input bitstream into independently decodable Group of Pictures (GOPs) for parallel processing, determining encoding parameters for coding tree units (CTUs) using extracted information, and encoding these CTUs into GOP-H.265/HEVC bitstreams, which are then combined into a final HEVC bitstream, utilizing multi-level architecture and wavefront parallel processing to expedite the transcoding process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If HEVC encoding is implemented on general-purpose processors, then compression efficiency is improved by 50%, but computational complexity becomes extremely high making it difficult to implement in software
Solution Approach 1:
The patent segments the video encoding process by dividing the bitstream into independently decodable Group of Pictures (GOPs) that can be processed in parallel. Each GOP is further divided into coding tree units (CTUs) that are processed independently using wavefront parallel processing, reducing the computational burden on single processors while maintaining HEVC compression efficiency.
Solution Approach 2:
The patent performs preliminary actions by extracting encoding information from the previously encoded H.264 bitstream before actual HEVC encoding. This includes parsing GOP structures, identifying CTU boundaries, and pre-processing motion compensation data, which reduces the computational complexity during the actual encoding phase.
2Manufacturing precision
If traditional HEVC encoding is used, then compression performance is superior, but processing speed is too slow for real-time systems
Solution Approach 1:
The patent divides the video stream into multiple independently decodable GOPs that can be encoded in parallel. Each GOP is further segmented into CTUs that are processed simultaneously using wavefront parallel processing, achieving real-time processing speeds while maintaining HEVC compression performance.
Solution Approach 2:
The patent introduces parallel processing across multiple dimensions: temporal dimension through independent GOP processing, spatial dimension through wavefront parallel processing of CTUs, and computational dimension through multi-threaded execution on multi-core processors, thereby achieving real-time processing without compromising compression performance.
3Manufacturing precision
If legacy H.264 content is transcoded to HEVC, then compression efficiency is improved, but transcoding time is extremely long
Solution Approach 1:
The patent performs preliminary actions by parsing the H.264 bitstream to extract GOP structures, motion compensation information, and other encoding parameters before actual HEVC encoding. This pre-processing reduces the transcoding time while maintaining the compression efficiency benefits of HEVC.
Solution Approach 2:
The patent uses an intermediary approach by leveraging the existing H.264 encoding information as a bridge to facilitate faster HEVC encoding. The extracted H.264 motion vectors and partitioning information serve as intermediaries that reduce the computational burden during HEVC transcoding, significantly reducing transcoding time.
Data Source
AI summary
The latest High Efficiency Video Coding (HEVC) standard achieves significant compression efficiency improvement over the H.264 standard, at a much higher cost of computational complexity. A framework for software-based H.264 to HEVC transcoding on multicore processors and distributed systems is provided. By utilizing information extracted from the input H.264 bitstream, the transcoding process can be accelerated at a high ratio with modest visual quality loss. Wavefront Parallel Processing (WPP) and SIMD acceleration are also implemented to improve the trans coding efficiency on multi-core processors with SIMD instruction set extensions. Based on the HEVC HM 12.0 reference software and using standard HEVC test bitstreams, the proposed transcoder can achieve around 120× speed up over decoding and re-encoding based on ffmpeg and the HM software without significant R-D performance loss.