Parallel Table-Based Bit Rate Estimator for HEVC
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing hardware architectures for HEVC video encoding face challenges in achieving high-throughput table-based CABAC rate estimation due to serial processing of syntax elements, leading to long computation times and low throughput, especially in real-time applications like live television and autonomous vehicle control.
Innovation Solution
A highly parallel hardware architecture is implemented, where syntax elements are classified into independent groups and processed in parallel, using local context tables and a global context model to facilitate simultaneous binarization and rate estimation, reducing hardware complexity and increasing accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If serial processing of syntax elements is used in traditional CABAC rate estimation, then hardware complexity is reduced, but throughput becomes low and computation time becomes long
Solution Approach 1:
The patent divides syntax elements into multiple independent groups (e.g., luma syntax elements, chroma syntax elements, and other syntax elements) that can be processed in parallel. Each group is assigned its own context model table and processed by separate hardware units, enabling simultaneous processing of multiple syntax elements without sequential dependencies. This segmentation resolves the contradiction by maintaining manageable hardware complexity through structured division while achieving high throughput through parallel execution of multiple groups.
2Loss of time
If parallel processing of syntax elements is implemented to increase throughput, then computation time is reduced, but hardware complexity increases
Solution Approach 1:
The patent assigns different context model tables to different syntax element groups based on their specific characteristics and probability distributions. Each parallel processing unit uses its own specialized context model optimized for its group, rather than sharing a single generic context model. This local quality approach enables efficient parallel processing with reduced computation time while controlling hardware complexity through targeted specialization rather than uniform duplication of all resources.
3Measurement precision
If context-adaptive binary arithmetic coding (CABAC) is used for rate estimation, then rate estimation accuracy is improved, but processing becomes extremely time-consuming
Solution Approach 1:
The patent pre-computes and stores context model probability values in lookup tables during an initialization phase, before actual rate estimation processing begins. During runtime, the parallel processing units simply retrieve pre-computed probability values from these tables rather than computing them on-the-fly. This preliminary action maintains the accuracy benefits of context-adaptive modeling while dramatically reducing processing time through efficient table lookups in parallel hardware units.
Data Source
AI summary
A highly parallel bit rate estimator and method for estimating bit rate are provided for high efficiency video coding applications. The electrical hardware architecture groups syntax elements into independent processing groups and utilizes a table-based context-adaptive binary arithmetic coding scheme to more rapidly estimate bit rates and optimize compression of high-definition videos.