Optical Flow Prediction Refinement for Image Decoding Efficiency
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 or bits, necessitating high-efficient image compression technology.
Innovation Solution
An image encoding/decoding method and apparatus that utilizes bi-directional optical flow (BDOF) or prediction refinement with optical flow (PROF) based on determining whether a current subpicture is treated as a picture, fetching prediction samples, and deriving refined prediction samples within specified ranges, with flag information signaled through a sequence parameter set (SPS).
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 extracts and transmits only the essential visual information by applying BDOF and PROF techniques that refine prediction samples to capture only the most important deviations from reference pictures, thereby maintaining high image quality while reducing the total amount of data that needs to be transmitted and stored
Solution Approach 2:
The patent changes the parameter of information representation by using optical flow-based prediction refinement instead of traditional block-based methods, allowing for more efficient encoding that maintains visual fidelity while reducing bitstream volume through smarter parameter selection and transmission
2Measurement precision
If BDOF or PROF is applied to current block, then prediction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent performs preliminary action by fetching prediction samples from reference pictures before applying BDOF or PROF refinement, and uses flag information to pre-determine whether refinement should be applied to each subpicture, thereby reducing real-time computational complexity while maintaining prediction accuracy
Solution Approach 2:
The patent applies partial action by selectively applying BDOF or PROF only to certain current blocks based on flag information indicating whether the current subpicture is treated as a picture, rather than applying the computationally intensive refinement to all blocks, thus balancing prediction accuracy with computational complexity
Data Source
AI summary
An image decoding method performed by an image decoding apparatus includes: determining that bi-directional optical flow (BDOF) or prediction refinement with optical flow (PROF) is applied to a current block; generating a prediction sample of the current block from a reference picture of the current block based on motion information of the current block; and deriving a refined prediction sample for the current block, by applying BDOF or PROF to the generated prediction sample of the current block, where the generating the prediction sample of the current block is performed based on whether a current subpicture including the current block is treated as a picture, where the generating the prediction sample of the current block comprises obtaining a reference sample specified by a position inside the reference picture and left-shifting a value of the reference sample, where the position inside the reference picture is clipped in a predetermined range.