Video Coding Mode Selection via Probability Partitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for mode selection in video encoders, particularly Wyner-Ziv encoders, face high computational complexity and fail to account for the global probability distribution, making them inefficient for vector-based encoding.
Innovation Solution
A method that transforms video frames into coefficients, partitions them based on probability statistics, computes associated probability distributions, and iteratively revises the partition vector to select encoding modes, reducing computational complexity while accounting for global probability distributions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If multiple encoder runs are employed for each block using available encoding modes, then compression performance is optimized, but computational complexity becomes exponentially high for vector-based encoding
Solution Approach 1:
The patent divides the video frame into multiple blocks and processes each block independently through multiple encoder runs. This segmentation allows the system to achieve good compression performance for each block without requiring exponential computational resources across the entire frame, as each block's mode selection is handled separately rather than evaluating all possible mode combinations globally.
Solution Approach 2:
The patent performs multiple encoder runs for each block to evaluate different encoding modes, but limits the scope to individual blocks rather than the entire video frame. This partial action approach achieves sufficient compression optimization at the block level without the excessive computational cost of global optimization, balancing performance and complexity.
2Device complexity
If conventional mode selection methods are used that evaluate mode-dependent functions on each block, then computational complexity is reduced, but global probability distribution is ignored leading to suboptimal compression
Solution Approach 1:
The patent performs preliminary encoding passes for each block to gather statistics about coefficient distributions before final mode selection. This preliminary action allows the system to capture global probability distribution characteristics without requiring a full exhaustive search, enabling more informed mode decisions with reduced computational overhead compared to conventional single-pass methods.
Solution Approach 2:
The patent uses feedback from multiple encoder runs on each block to refine mode selection decisions. By evaluating mode-dependent functions across multiple passes and using the results to inform subsequent encoding decisions, the system achieves better compression efficiency while maintaining manageable computational complexity through iterative refinement rather than exhaustive search.
3Manufacturing precision
If exhaustive mode evaluation is performed for Wyner-Ziv vector-based encoding, then optimal compression is achieved, but computational complexity becomes infeasible
Solution Approach 1:
The patent segments the Wyner-Ziv encoding process into independent block-level operations, performing mode evaluation and selection for each block separately. This segmentation makes the encoding process feasible by avoiding the need to evaluate all possible mode combinations for the entire vector at once, thus maintaining encoding speed while still achieving optimization at the block level.
Solution Approach 2:
The patent performs a limited number of encoder runs for mode evaluation purposes rather than exhaustive evaluation of all possible modes. This partial action approach achieves sufficient compression optimization for practical applications without the prohibitive computational cost of complete exhaustion, balancing encoding speed with compression quality.
Data Source
AI summary
A method and system are disclosed for selecting a mode to encode video data. The method comprises the steps of (a) transforming a source video frame into a set of coefficients, (b) partitioning said set of coefficients into a plurality of subsets of the coefficients on the basis of probability statistics corresponding to a plurality of encoding modes, wherein each of said subsets is identified for encoding by one of the plurality of encoding modes. The method comprises the further steps of (c) for each of the plurality of subsets of coefficients, computing defined parameters of an associated probability distribution for said subset, and (d) repeating steps (b) and (c) until a predetermined termination condition is satisfied. When this predetermined termination condition is satisfied, the subsets of coefficients, as they exist at that time, are output to a video encoder, which preferably is a Wyner-Ziv encoder.