Non-Causal Overlapped Block Prediction in Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face inefficiencies in processing and transmitting digital video streams due to high data volume, particularly in applications like video conferencing, where non-causal overlapped block prediction is not permitted, limiting the use of temporal and spatial similarities across frames.
Innovation Solution
The method involves obtaining prediction parameters for adjacent blocks, generating a base prediction for a current block, identifying an overlap region, and creating an overlapped prediction for each pixel within that region as a function of the base and adjacent predictions, allowing for non-causal overlapped block prediction modes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If non-causal overlapped block prediction is used, then coding efficiency is improved, but causality constraints in traditional video coding are violated
Solution Approach 1:
The current block is divided into two distinct regions: an overlap region that utilizes non-causal prediction from future blocks, and a non-overlap region that uses traditional causal prediction. This segmentation allows the patent to apply advanced prediction techniques where beneficial while maintaining compatibility with standard video coding constraints in other areas.
Solution Approach 2:
Different prediction strategies are applied to different spatial regions within the current block. The overlap region employs non-causal overlapped block prediction for improved accuracy, while the non-overlap region uses conventional prediction methods. This local differentiation optimizes overall coding efficiency without requiring non-causal prediction throughout the entire block.
2Quantity of substance
If data volume in video streams is reduced, then transmission and storage requirements are lowered, but prediction accuracy may deteriorate
Solution Approach 1:
Prediction parameters for all adjacent blocks (including future blocks in non-causal prediction) are obtained in advance before encoding the current block. This preliminary acquisition of prediction data from multiple directions enables the construction of accurate overlapped predictions without increasing the final bitstream volume, as the prediction process occurs during encoding rather than requiring additional transmitted data.
Solution Approach 2:
The prediction for the overlap region is formed by combining multiple prediction sources: the base prediction from the current block's motion compensation and adjacent predictions from surrounding blocks. This composite prediction approach leverages spatial and temporal correlations from multiple directions to achieve higher accuracy than single-source prediction, thereby reducing residual data volume.
3Productivity
If processing efficiency is improved through non-causal prediction, then encoding speed increases, but computational complexity for obtaining all adjacent block parameters increases
Solution Approach 1:
Non-causal overlapped block prediction is applied selectively only to the overlap region of the current block, rather than to the entire block. The non-overlap region uses standard prediction methods. This partial application reduces the computational overhead of non-causal prediction while still achieving efficiency gains in the regions where it provides the most benefit.
Solution Approach 2:
The patent merges the base prediction process with the overlapped prediction process by integrating the acquisition of adjacent block parameters into the existing encoding workflow. By combining these operations and reusing previously computed prediction parameters where possible, the patent reduces redundant computations while maintaining the efficiency benefits of non-causal prediction.
Data Source
AI summary
A method for processing a selected portion of a video, the selected portion of the video having a plurality of blocks. The method includes obtaining current prediction parameters for all of a plurality of adjacent blocks from the plurality of blocks that are adjacent to a current block from the plurality of blocks in the selected portion of the video, generating a base prediction for the current block from the plurality of blocks using the current prediction parameters associated with the current block, identifying adjacent prediction parameters from the current prediction parameters for a first adjacent block from the plurality of adjacent blocks, determining an overlap region within the current block and adjacent to the first adjacent block, and generating, for each pixel within the overlap region, an overlapped prediction for the pixel as a function of the base prediction and a prediction based on the adjacent prediction parameters.