Hierarchical Video Coding Unit Splitting for Compression Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video codecs face challenges in efficiently encoding and decoding high-resolution video content due to limitations in encoding methods based on fixed macroblock sizes, which result in decreased data compression efficiency and increased complexity.
Innovation Solution
The proposed solution involves a hierarchical structure for encoding and decoding video using pattern information, where a picture is split into maximum coding units and smaller coding units, with transformation units determined based on depth levels, allowing for adaptive encoding and decoding modes to minimize encoding errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fixed macroblock size is used for encoding, then the encoding structure is simple, but the data compression efficiency decreases and encoding complexity increases for high-resolution video
Solution Approach 1:
The video picture is divided into multiple maximum coding units, which are further divided into smaller coding units at different depth levels. This hierarchical segmentation allows the encoder to adapt to high-resolution video by processing smaller regions independently, improving compression efficiency while maintaining manageable complexity through structured organization.
Solution Approach 2:
The coding unit size and structure are made dynamic through the depth parameter, which allows different regions of the picture to be encoded at different granularity levels. This dynamic adaptation enables the encoder to optimize for local image characteristics, improving overall compression efficiency for high-resolution content.
2Loss of information
If the maximum coding unit size is increased, then fewer coding units are needed reducing overhead, but the precision of texture representation decreases
Solution Approach 1:
By segmenting the picture into maximum coding units and further into smaller coding units at different depths, the system can represent fine texture details in regions that require it while using larger units in smoother regions, thus preserving texture information without unnecessarily increasing the total number of coding units.
Solution Approach 2:
Different regions of the picture can be encoded with different coding unit sizes and depths based on local texture characteristics. Regions with complex texture use smaller coding units at greater depths to preserve detail, while smooth regions use larger coding units, optimizing the balance between texture representation and the number of coding units.
3Adaptability or versatility
If hierarchical coding units with multiple depths are used, then encoding adaptability improves, but the encoding complexity increases
Solution Approach 1:
The hierarchical structure segments the encoding process into distinct depth levels, where each level processes coding units of a specific size. This segmentation provides adaptability to different video content types while controlling complexity by organizing the encoding process into manageable, repeatable stages across different depths.
Solution Approach 2:
The same encoding and decoding processes are applied universally across all depth levels and coding units. This multi-functionality approach allows the system to handle diverse video content adaptively while reusing the same core algorithms, thereby improving versatility without proportionally increasing complexity.
4Measurement precision
If more pattern information is encoded, then texture representation accuracy improves, but the bitstream size increases
Solution Approach 1:
Pattern information and texture details are encoded with higher precision only in regions where it is necessary, determined by local image characteristics and coding unit depth. This selective encoding improves texture representation accuracy where needed while minimizing the overall bitstream size by using coarser representation in less critical regions.
Solution Approach 2:
The amount of pattern information encoded is dynamically adjusted based on the coding unit depth and local texture complexity. Deeper levels with smaller coding units encode more detailed pattern information, while shallower levels use compressed representations, optimizing the balance between accuracy and bitstream size.
Data Source
AI summary
A video decoding apparatus including an extractor which extracts from a bitstream first pattern information indicating whether residual samples of a current coding unit are equal to 0, and when the first pattern information indicates the residual samples are not equal to 0, extracts from the bitstream transformation index information indicating whether a transformation unit of a current level included in the current coding unit is split, a decoder which splits the transformation unit of the current level into transformation units of a lower level when the transformation index information indicates a split of the transformation unit of the current level, wherein the extractor further extracts second pattern information for the transformation unit of the current level when the transformation index information indicates a non-split of the transformation unit of the current level, wherein the second pattern information indicates whether the transformation unit of the current level contains one or more transform coefficients not equal to 0, wherein the transformation unit of the current level is split into four transformation units of the lower level.