Video Encoding Emulating Reduced Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression methods, such as those using Reference Picture Resampling (RPR), face limitations when the reference picture and current picture are not at the same spatial resolution, leading to suboptimal compression efficiency and visual artifacts.
Innovation Solution
A method and apparatus for video encoding and decoding that adapt to emulate a reduced resolution for parts of the picture by signaling specific syntax data elements, allowing processes like transform, motion compensation, and filtering to operate as if the resolution was changed, without actually scaling the coded frame, thereby maintaining coding efficiency and visual quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If Reference Picture Resampling (RPR) is used to dynamically re-scale pictures, then compression efficiency is improved, but device complexity increases and some coding tools become unavailable due to resolution mismatch between reference and current pictures
Solution Approach 1:
The patent creates a virtual copy of the reference picture at the original resolution by emulating the inverse of the resampling operation. Instead of actually rescaling the reference picture, the decoder synthesizes a virtual reference picture that appears to be at the original resolution, allowing all coding tools to operate as if no resampling occurred. This copying approach maintains compression efficiency while avoiding the complexity of actual resampling operations.
Solution Approach 2:
The patent introduces an intermediary virtual reference picture that mediates between the downsampled reference picture and the current picture. This virtual reference picture serves as an intermediate representation that allows motion compensation and other coding tools to function correctly without requiring actual resolution changes. The intermediary enables the system to benefit from RPR compression while maintaining tool availability.
2Productivity
If Reference Picture Resampling (RPR) is applied to achieve better rate-quality trade-off, then compression performance is improved, but visual artifacts increase due to resolution mismatch
Solution Approach 1:
Instead of directly applying resampling to the reference picture, the patent inverts the approach by emulating the inverse operation. The decoder synthesizes a virtual reference picture that represents what the reference picture would look like at the original resolution, thereby avoiding the visual artifacts that would result from actual resampling. This inversion strategy maintains compression performance while eliminating harmful visual artifacts.
3Productivity
If actual resolution scaling is performed on reference pictures, then coding efficiency is improved for certain tools, but the availability of coding tools is reduced due to resolution changes
Solution Approach 1:
The patent makes the virtual reference picture universal by ensuring it can serve all coding tools regardless of their resolution requirements. By emulating the original resolution through virtual synthesis, the system enables all coding tools to function with the reference picture, achieving multi-functionality. This universality allows the system to maintain high coding efficiency while preserving tool availability, as all tools can operate on the virtual reference picture as if it were at the original resolution.
Data Source
AI summary
At least a method and an apparatus are presented for efficiently encoding or decoding video. For example, video data representative of a part of a picture is obtained, at least one syntax data element specifying to adapt the encoding/decoding to emulate a reduced resolution applied to the part of the picture is signaled; and video data is encoded/decoded wherein at least one process of the encoding/decoding is adapted to emulate a reduced resolution applied to the part of the picture.