Temporal Prediction in Video Coding Using Block Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video coding methods are limited in coding efficiency and introduce blocking artifacts due to insufficient use of prediction data in reference frames, with existing transformation and motion compensation techniques failing to adequately describe differences between current and reference frames.
Innovation Solution
A method that combines motion information and non-motion related transformation models across multiple image blocks to determine a prediction for the block being coded, using conditional motion-compensated prediction and filtering to reduce blocking artifacts and enhance coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If block-level transformations of reference data are applied to improve prediction quality, then coding efficiency is improved, but overhead increases and blocking artifacts are accentuated
Solution Approach 1:
The patent merges block-level transformation parameters with motion compensation to create a unified prediction process. By combining the transformation operation with motion compensation in a single integrated algorithm, the system achieves improved prediction quality without requiring separate block-level transformation processing, thus reducing overhead while maintaining coding efficiency benefits.
Solution Approach 2:
The patent makes the motion compensation process universal by incorporating transformation operations within it. The motion compensation algorithm can handle both standard motion compensation cases and transformed reference data cases through a single unified processing path, eliminating the need for separate block-level transformation processing and reducing overall system overhead.
2Productivity
If conventional motion estimation and compensation are used, then processing is simple, but prediction accuracy is insufficient leading to limited coding gains
Solution Approach 1:
The patent changes the parameters of the motion compensation process by introducing transformation parameters (such as brightness, contrast, or other image quality parameters) that can be adjusted to optimize prediction accuracy. By allowing these parameters to be modified based on the specific characteristics of the video data and reference frames, the system achieves improved prediction accuracy and corresponding coding gains without fundamentally changing the motion compensation framework.
3Ease of manufacture
If independent block-level processing is used in conventional video coding, then processing is straightforward, but blocking artifacts are introduced and rate-distortion performance is limited
Solution Approach 1:
The patent merges the transformation operation with motion compensation to create a unified prediction process that operates at a higher level than individual blocks. This merging allows the system to maintain the simplicity of block-based processing while eliminating blocking artifacts through the coordinated action of transformation and motion compensation across multiple blocks.
Solution Approach 2:
The patent introduces transformation parameters as intermediaries between the reference frame data and the motion compensation process. These intermediaries (transformation parameters) mediate the interaction between reference data and current frame prediction, allowing smooth transitions and reducing blocking artifacts by acting as a buffer that reconciles differences between blocks.
Data Source
AI summary
A method, apparatus and article of manufacture for performing temporal prediction are described. In one embodiment, a decoding method includes, but is not limited to, generating a prediction for a sample in one of multiple blocks using transformation parameters associated with the one block and transformation parameters associated with at least one other block in the group of blocks and adding residual data to the prediction to obtain a decoded sample.