Inter-prediction Weighted Sum for Block Boundary Error Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-resolution and high-quality image data requires efficient compression techniques to reduce transmission and storage costs, as conventional methods struggle with increased data volume, leading to higher prediction errors at block boundaries during inter-prediction.
Innovation Solution
The method involves deriving prediction samples for a current block by calculating motion vectors and using a weighted sum of reference units from adjacent blocks to reduce prediction errors at block boundaries, enhancing inter-prediction performance and coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional inter-prediction methods are used for high-resolution images, then transmission and storage costs increase due to increased data volume, but prediction performance may be maintained with standard block-based approaches
Solution Approach 1:
The current block is divided into multiple subunits along the block boundary, allowing selective application of the weighted sum method only to boundary regions rather than the entire block. This segmentation reduces computational complexity while maintaining prediction accuracy where it matters most - at the block boundaries where prediction errors typically occur.
Solution Approach 2:
The patent applies different prediction methods to different regions: the weighted sum of reference units is applied specifically to subunits at the block boundary, while other regions use conventional prediction. This local quality approach optimizes prediction performance at critical boundary areas without unnecessarily increasing data volume or computational load for the entire block.
2Productivity
If block-based inter-prediction is used, then coding efficiency is improved through compression, but prediction errors occur at block boundaries
Solution Approach 1:
The patent modifies the prediction parameters at block boundaries by using a weighted sum of reference units from adjacent blocks instead of standard block-based prediction. This parameter change specifically targets boundary regions where prediction errors occur, improving prediction accuracy without affecting the overall coding efficiency of the block-based approach.
Solution Approach 2:
The patent introduces an intermediary calculation method - the weighted sum of reference units - that acts as a mediator between the current block and reference blocks at their boundaries. This intermediary approach smooths transitions at block boundaries by combining information from multiple reference sources, reducing prediction errors while maintaining compression efficiency.
3Measurement precision
If high-resolution images are transmitted with standard compression, then image quality is preserved, but transmission and storage costs increase significantly
Solution Approach 1:
The patent extracts and separately processes only the boundary regions of blocks using the weighted sum method, rather than applying complex processing to the entire high-resolution image. This extraction approach maintains image quality at critical boundaries while minimizing the amount of data that requires enhanced processing, thereby reducing overall transmission and storage costs.
Data Source
AI summary
An inter-prediction method according to the present invention comprises the steps of: deriving a first movement vector of the current block; deriving prediction samples for the current block on the basis of the first movement vector; deriving a subunit located in the current block and adjacent to the target boundary of same; deriving a first reference unit on the basis of the subunit and first movement vector; deriving a second movement vector of a neighboring block adjacent to the target boundary; deriving a second reference unit on the basis of the subunit and second movement vector; and deriving the modified prediction samples on the basis of weighted sum of the first reference unit and second reference unit. According to the present invention, by increasing the effectiveness of inter-prediction at the block boundary, amount of data required for residual data can be reduced and the overall coding efficiency can be increased.