Video Packetization for Parallel Decoding of Ultra-HD Streams
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Decoding ultra-high definition moving images such as 8K requires significant processing load, making real-time decoding challenging with a single decoder, especially due to high bit rates and complex multiplexing methods like MPEG-2 TS or MMT, which complicates demultiplexing and parallelization.
Innovation Solution
The method involves encoding a video signal into access units, dividing them into packet groups, and transmitting control information including presentation and decoding times, allowing for parallel decoding across multiple decoders by optimizing packetization and demultiplexing processes, particularly using MMT packetization to efficiently manage slice segments and reduce processing load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If encoded data is multiplexed using MPEG-2 TS or MMT methods, then transmission efficiency is improved, but demultiplexing complexity increases
Solution Approach 1:
The patent divides an access unit into multiple packet groups, where each packet group contains a specific portion of the access unit data. This segmentation allows the receiving device to process and demultiplex smaller, more manageable packet groups independently, reducing the overall demultiplexing complexity while maintaining transmission efficiency through structured organization of encoded data
2Device complexity
If a single decoder is used for ultra-high definition moving images, then device complexity is reduced, but real-time decoding capability deteriorates
Solution Approach 1:
The patent enables parallel decoding by dividing the encoded data into multiple packet groups that can be decoded simultaneously by multiple decoders. Each decoder processes a specific packet group independently, allowing ultra-high definition moving images to be decoded in real-time through parallel processing while keeping individual decoder complexity manageable
Solution Approach 2:
The patent introduces temporal dimension to the decoding process by using multiple decoders operating in parallel across different time streams. This dimensional approach transforms a single sequential decoding bottleneck into multiple concurrent decoding paths, achieving real-time processing capability for ultra-high definition content
3Device complexity
If access units are packetized as single units, then packet structure simplicity is improved, but parallel processing capability deteriorates
Solution Approach 1:
The patent segments each access unit into multiple packet groups, where each packet group contains a specific portion of the access unit data along with appropriate control information. This segmentation structure enables receiving devices to distribute different packet groups to multiple decoders for parallel processing, while maintaining relatively simple packet structures that are easy to parse and process
4Productivity
If multiple decoders are used for parallel decoding, then decoding speed is improved, but synchronization difficulty increases
Solution Approach 1:
The patent incorporates presentation time and decoding time information in advance within each packet group's control information. This preliminary timing information allows multiple decoders to synchronize their processing automatically without requiring complex external synchronization mechanisms, as each decoder can independently determine when to process and output its decoded data based on the pre-provided timing parameters
Data Source
AI summary
A transmitting method according to one aspect of the present disclosure includes: encoding a video signal and generating encoded data including a plurality of access units; storing the plurality of access units in a packet in a unit that defines one access unit as one unit or in a unit defined by dividing one access unit, and generating a packet group; transmitting the generated packet group as data; generating first information and second information, the first information indicating a presentation time of a first access unit that is presented first among the plurality of access units, and the second information being used to calculate a decoding time of the plurality of access units; and transmitting the first information and the second information as control information.