Intra-prediction Mode Reordering for Video Encoding Memory Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital video encoding techniques require large data tables for prediction modes, leading to high memory costs and increased complexity, which is burdensome for low-cost machines and inefficient in terms of bandwidth and storage.
Innovation Solution
The proposed solution involves reordering prediction modes based on their likelihood of producing a lesser prediction error, using a systematic approach to arrange horizontal, vertical, diagonal, and intermediate angle prediction modes, and employing intraframe coding to exploit spatial redundancies within video images, thereby reducing memory requirements and improving coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If large data tables are used for prediction modes, then prediction accuracy is improved, but memory requirements and device complexity increase
Solution Approach 1:
The patent segments the prediction mode data into multiple groups or categories, allowing the system to manage and process prediction modes in smaller, more manageable units rather than requiring a single large data table, thus reducing memory requirements while maintaining prediction accuracy
Solution Approach 2:
The patent changes parameters such as the organization structure, ordering, or representation of prediction mode data to enable more efficient storage and processing, reducing the memory footprint and device complexity while preserving the essential prediction functionality
2Measurement precision
If large data tables are used for prediction modes, then prediction accuracy is improved, but memory cost increases
Solution Approach 1:
The patent extracts only the essential or most frequently used prediction mode data from a complete large data table, storing only the necessary subset in memory while maintaining access to full prediction accuracy when needed, thereby reducing memory requirements
Solution Approach 2:
The patent applies parameter changes to optimize the storage representation of prediction mode data, such as using compressed formats, efficient data structures, or selective storage strategies that reduce memory requirements while preserving prediction accuracy
3Productivity
If traditional intraframe encoding is used, then spatial redundancies are exploited, but bandwidth and storage requirements remain high
Solution Approach 1:
The patent applies preliminary ordering and organization of prediction modes before the actual encoding process, preparing the data in advance to maximize the exploitation of spatial redundancies and enable more efficient compression, thereby reducing bandwidth and storage requirements
Solution Approach 2:
The patent changes parameters related to the organization and sequencing of prediction modes to optimize compression efficiency, enabling better exploitation of spatial redundancies and achieving higher coding efficiency with reduced bandwidth and storage requirements
Data Source
AI summary
Embodiments of the present invention relate to methods and systems for ordering, communicating and applying pixel intra-prediction modes.