Image Bitstream Handling for Neural Post-Filter Picture Sync
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images leads to a significant increase in transmission and storage costs due to the rise in the amount of transmitted information, necessitating high-efficient image compression technologies that can effectively process neural-network post-filters without errors and prevent mismatches between encoder and decoder input picture lists.
Innovation Solution
An image encoding/decoding method and apparatus that restricts the inclusion of neural-network post-filter-related SEI messages to non-discardable and non-output pictures, ensuring efficient encoding/decoding and preventing mismatches by managing input pictures effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If high-resolution and high-quality images are transmitted, then image quality is improved, but transmission cost and storage cost increase
Solution Approach 1:
The patent changes the parameter of picture type classification by introducing a new criterion: whether a picture is used as input for neural-network post-filter. This creates a new category (pictures not used as NNPF input) that can be discarded, allowing the system to maintain high image quality for important pictures while reducing transmission and storage of less important ones, thus resolving the contradiction between image quality and data volume
2Manufacturing precision
If neural-network post-filter is applied, then image quality is improved, but mismatch between encoder and decoder input picture lists occurs
Solution Approach 1:
The patent applies preliminary action by having the encoder proactively determine which pictures will be used as input for neural-network post-filter and communicate this information to the decoder in advance. This allows the decoder to prepare its input picture list accordingly, ensuring both encoder and decoder have synchronized understanding of which pictures to process, thus preventing mismatches while enabling quality improvement through NNPF
3Loss of information
If all pictures are included in encoding, then completeness of data is improved, but processing efficiency decreases
Solution Approach 1:
The patent extracts and separates pictures that are not used as input for neural-network post-filter from the complete picture set. By identifying and extracting this specific subset, the system can discard these pictures without losing important information (since they're not needed for NNPF), thereby improving encoding efficiency while maintaining completeness of essential picture data
Data Source
AI summary
An image encoding/decoding method, a method of transmitting a bitstream and a computer-readable recording medium storing a bitstream are provided. An image decoding method may comprise obtaining unit information including a current picture and coding the current picture based on the unit information. A type of the current picture may be restricted based on a neural-network post-filter (NNPF)-related supplemental enhancement information (SEI) message included in the unit information.