Motion Field Processing for Hardware-Friendly Video Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding standards fail to recognize the quality properties of candidate reference frames, leading to sub-optimal motion vector prediction and inefficient coding performance, and lack hardware-friendly designs for motion field processing, which complicates hardware implementation and reduces parallel processing efficiency.
Innovation Solution
A method to select reference frames for motion vector prediction based on temporal distance, coding quality, and hierarchical relationships, combined with adaptive motion projection order and hole derivation techniques using processing blocks of predetermined sizes, to improve coding efficiency and reduce hardware complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing video coding standards are used for motion vector prediction, then implementation is simple, but coding efficiency is sub-optimal due to failure to recognize quality properties of candidate reference frames
Solution Approach 1:
The patent changes the selection parameters for reference frames from simple temporal distance to a composite quality metric that includes temporal distance, coding quality, and hierarchical relationships. This allows the system to prioritize high-quality reference frames for motion vector prediction, improving coding efficiency while maintaining manageable complexity through structured quality assessment
Solution Approach 2:
The patent performs preliminary quality assessment and hierarchical organization of reference frames before motion vector prediction. By pre-evaluating and ranking candidate reference frames based on multiple quality factors, the system prepares optimized prediction lists in advance, improving coding efficiency without adding complexity during the actual prediction process
2Productivity
If adaptive motion projection order is implemented, then coding efficiency improves, but hardware implementation complexity increases
Solution Approach 1:
The patent segments the motion field processing into discrete processing blocks of predetermined sizes. Each block can be processed independently with adaptive motion projection, allowing hardware to handle complexity in modular units rather than as a monolithic system. This segmentation enables parallel processing while maintaining adaptive quality
Solution Approach 2:
The patent implements dynamic motion projection order that adapts to local characteristics of each processing block. Rather than applying a fixed projection order globally, the system dynamically adjusts the projection strategy based on motion complexity and quality requirements of each block, improving coding efficiency while allowing hardware to optimize processing paths
3Manufacturing precision
If hole derivation techniques are used, then motion field quality improves, but hardware processing complexity increases
Solution Approach 1:
The patent applies hole derivation techniques selectively to specific regions where motion field quality needs improvement, rather than uniformly across the entire frame. By identifying areas with poor prediction quality and applying localized hole filling using surrounding block information, the system improves motion field quality where needed while minimizing unnecessary processing elsewhere
Solution Approach 2:
The patent uses copying techniques to fill holes in the motion field by replicating motion vector information from surrounding processed blocks. Instead of performing complex calculations for each missing value, the system copies and interpolates from neighboring blocks, improving motion field completeness with simpler hardware operations
Data Source
AI summary
Techniques are proposed to improve temporal motion projection in video coding. Candidate reference frames available for use in temporal motion projection are sorted in processing order according to scores assigned based on estimates of the reference frames' suitability for prediction. Such estimates may be based on temporal distance between each candidate reference frame and that reference frame's prediction references. Estimates may be based, for each reference frame, based on an estimate of coding quality of a reference frame from which the respective candidate reference frame makes a prediction reference. Once sorted, the candidate reference frames may be processing in the sorting order to supply prediction data to a current frame that is to be coded from the candidate reference frames. Additionally, hardware friendly designs of motion field hole filling and motion vector smoothing operations are proposed. Such designs can reduce hardware implementation complexity and benefit hardware parallel processing in several aspects: by removing the dependency among different processing block rows for hole filling and motion vector smoothing so that it is becomes easier and friendlier to achieve hardware parallel processing; by reducing the hardware bandwidth loading overhead; by improving hardware pipeline throughput; and/or by avoiding adding a line buffer to store the data from the above row since a line buffer will increase hardware cost.