Sub-picture Raster Scanning for UHD Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high memory bandwidth requirements for motion estimation in ultra-high definition (UHD) video coding pose a significant challenge for cost-effective chip design, as UHD video requires larger search ranges and exceeds the capabilities of single-core processors.
Innovation Solution
The method involves dividing an image into sub-pictures and encoding them in a sub-picture based raster scanning order, allowing for parallel processing on multi-cores. This approach reduces on-chip memory requirements and enables efficient multi-core processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If traditional raster scanning order is used for UHD video coding, then coding efficiency can be maintained, but on-chip memory requirements become excessively large
Solution Approach 1:
The patent divides the traditional raster scanning order into sub-picture based segments. Each sub-picture is processed independently with its own scanning order, allowing parallel processing across multiple cores while reducing the memory footprint required for motion estimation and compensation in each sub-picture.
Solution Approach 2:
The patent introduces a new dimension to the scanning order by processing sub-pictures in parallel across multiple cores rather than sequentially. This dimensional change from sequential to parallel processing reduces the cumulative memory requirements while maintaining overall coding efficiency.
2Productivity
If single-core processing is used for UHD video coding, then device complexity is reduced, but processing capability becomes insufficient
Solution Approach 1:
The patent segments the video coding task across multiple cores by dividing the picture into sub-pictures. Each core processes a specific sub-picture independently, enabling parallel processing that increases overall capability while keeping each individual core's complexity manageable.
Solution Approach 2:
The patent creates a universal processing model where multiple identical cores can work in parallel on different sub-pictures. This multi-functional approach allows the system to scale processing capability by adding more cores without increasing the complexity of individual core design.
3Productivity
If larger search range is used for motion estimation in UHD, then coding efficiency improves, but on-chip memory requirements increase
Solution Approach 1:
The patent applies segmentation to motion estimation by processing each sub-picture with its own localized search range. This allows each core to use a manageable search range for its sub-picture while the collective effect provides the comprehensive search capability needed for UHD coding efficiency.
Data Source
AI summary
A method and apparatus for sub-picture based raster scanning coding order. The method includes dividing an image into even sub-pictures, and encoding parallel sub-pictures on multi-cores in raster scanning order within sub-pictures, wherein from core to core, coding of the sub-picture is independent around sub-picture boundaries, and wherein within a core, coding of a sub-picture is at least one of dependent or independent around sub-picture boundaries.