MPEG Decoder Parallel Processing via Macroblock Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
MPEG decoders and encoders face challenges in implementing parallel processing due to dependencies in variable-length decoding, peripheral prediction, and inter-frame prediction, limiting their processing performance.
Innovation Solution
The implementation of a moving picture decoding and encoding apparatus with variable-length decoding and image decoding/encoding units, along with a data buffer, allows for parallel processing at the macroblock line or slice level, synchronizing the operations of these units to enhance processing performance without increasing the overall operating frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If parallel processing is implemented in MPEG decoders, then processing performance is improved, but the complexity of maintaining processing sequence increases due to dependencies between variable-length decoding, peripheral prediction, and inter-frame prediction
Solution Approach 1:
The patent divides the MPEG decoding process into independent macroblock-level units that can be processed in parallel. Each macroblock is treated as an independent unit with its own variable-length decoding, peripheral prediction, and inter-frame prediction operations, eliminating the need to maintain a global processing sequence while improving overall performance.
Solution Approach 2:
The patent introduces a buffer as an intermediary component between the variable-length decoding unit and the image decoding unit. This buffer stores decoded macroblock data and provides synchronized access to multiple image decoding units, mediating the data flow and enabling parallel processing without compromising the integrity of the decoding sequence.
2Productivity
If variable-length decoding is performed in serial manner at high speed, then processing performance is improved, but the operating frequency needs to be increased
Solution Approach 1:
The patent segments the variable-length decoding process into multiple independent units that can operate in parallel. Instead of serial processing requiring high frequency, multiple lower-frequency units work simultaneously on different macroblocks, achieving the same overall throughput with reduced individual unit frequency requirements.
Solution Approach 2:
The patent transitions from temporal parallelism (serial processing at high frequency) to spatial parallelism (multiple simultaneous decoding units). This dimensional change allows the system to achieve high productivity without proportionally increasing the operating frequency of individual components.
3Productivity
If slice-based division is used for parallel processing, then processing performance is improved, but adaptability to different MPEG versions is reduced since slice-based division is not implemented in MPEG4 or H.264
Solution Approach 1:
The patent adopts macroblock-level segmentation instead of slice-based division. This segmentation approach is universally applicable across MPEG-2, MPEG-4, and H.264 standards since the macroblock is a common fundamental unit in all these formats, thereby maintaining adaptability while enabling parallel processing.
Solution Approach 2:
The patent creates a universal parallel processing framework based on macroblock independence that can function across multiple MPEG versions. The same macroblock-level parallel architecture works for MPEG-2, MPEG-4, and H.264, providing multi-functionality and broad adaptability without requiring version-specific implementations.
Data Source
AI summary
A moving picture decoding apparatus includes one or more variable-length decoding units, a data buffer configured to store data output from the one or more variable-length decoding units, and a plurality of image decoding units configured to read the data from the data buffer and to perform image decoding the data.