Shared Coded Picture Bitstream Switching for Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for viewport-dependent omnidirectional video streaming, such as SHVC ROI encoding, face limitations due to restricted decoding hardware support for inter-layer prediction, which hinders bitrate reduction and compression gains.
Innovation Solution
An enhanced encoding method is introduced, where an input picture sequence is encoded into two bitstreams with a shared coded picture, allowing switching between them, and indicating this switching in the media presentation description, enabling flexible decoding and bitrate adjustment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If SHVC ROI encoding with inter-layer prediction is used, then compression gain and bitrate reduction are improved, but hardware support limitations reduce reliability and usability
Solution Approach 1:
The video stream is segmented into multiple bitstreams with different quality levels. The method creates a first bitstream with first coded pictures and a second bitstream with second coded pictures, allowing the system to select and switch between different segments based on network conditions and device capabilities, thereby achieving bitrate reduction without relying on advanced hardware support for inter-layer prediction
Solution Approach 2:
A shared coded picture serves as an intermediary between the first and second bitstreams. This shared picture acts as a common reference that both bitstreams can utilize, enabling flexible switching while maintaining decoding compatibility across different hardware platforms without requiring full inter-layer prediction support
2Adaptability or versatility
If multiple bitstreams with shared coded pictures are used, then adaptability and flexibility are improved, but device complexity increases
Solution Approach 1:
The method merges the first and second bitstreams by introducing a shared coded picture that is common to both streams. This merging approach allows the encoder to manage multiple bitstreams with reduced complexity, as the shared picture eliminates redundant encoding operations and enables straightforward switching mechanisms based on simple identification information
Data Source
AI summary
A method comprising: encoding an input picture sequence into a first bitstream (1000), said encoding comprising: encoding a first input picture into first two coded pictures of the first bitstream, wherein one of said two coded pictures is a first shared coded picture (1000a); encoding a first set of input pictures into a first set of coded pictures of the first bitstream, following in decoding order the first two coded pictures of the first bitstream, wherein any picture subsequent to the first shared coded picture in decoding order is not predicted from any picture that precedes the first shared coded picture in decoding order and is not a shared coded picture (1000b); and encoding the input picture sequence into a second bitstream (1002), said encoding comprising: encoding the first input picture into first two coded pictures of the second bitstream, wherein one of said two coded pictures is the first shared coded picture (1002a); encoding the first set of input pictures into a second set of coded pictures of the second bitstream, following in decoding order the first two coded pictures of the second bitstream and differing from the first set of coded pictures of the first bitstream, wherein any picture subsequent to the first shared coded picture in decoding order is not predicted from any picture that precedes the first shared coded picture in decoding order and is not a shared coded picture (1002b).