Adaptive Macroblock Prediction for Bit-Efficient Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video signal processing methods struggle with inefficient compression and require excessive bits for coding regions, particularly when dealing with macroblocks larger than 16x16 pixels, leading to suboptimal coding efficiency and accuracy.
Innovation Solution
Adaptive use of macroblock types and prediction modes for macroblocks larger than 16x16, determining prediction block sizes based on transform size information, and deriving intra and inter prediction modes from neighboring blocks to reduce the number of bits required for coding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If macroblock type and prediction mode information is transmitted separately for each macroblock, then prediction accuracy is improved, but the number of bits required for coding increases
Solution Approach 1:
The macroblock is divided into sub-macroblocks, and prediction mode information is coded at the sub-macroblock level rather than the entire macroblock level. This segmentation allows for more precise prediction control in different regions while reducing the total number of bits needed compared to coding every parameter separately for the whole macroblock.
Solution Approach 2:
Multiple macroblocks are grouped into a macroblock group, and the same macroblock type and prediction mode information is applied to all macroblocks within the group. This merging strategy reduces the number of bits required by transmitting the information once for the group rather than separately for each macroblock, while still maintaining acceptable prediction accuracy through the grouping approach.
2Ease of manufacture
If fixed 16x16 macroblock size is used, then coding simplicity is maintained, but coding efficiency decreases for high-resolution video
Solution Approach 1:
The macroblock structure is made dynamic by allowing different macroblock types (including larger than 16x16) and configurable macroblock group sizes. This dynamic adaptability enables the system to optimize for high-resolution video content by using larger macroblocks when appropriate, thereby improving coding efficiency without completely sacrificing the simplicity of the fixed 16x16 approach.
Solution Approach 2:
The macroblock size parameter is changed from a fixed 16x16 to allow larger sizes (greater than 16x16 pixels). This parameter change enables better adaptation to high-resolution video content, improving coding efficiency by matching the macroblock size more closely to the actual content characteristics in high-resolution scenarios.
3Measurement precision
If intra prediction is performed for all macroblocks, then prediction accuracy is improved, but the number of bits required for coding increases
Solution Approach 1:
Intra prediction is applied selectively rather than uniformly to all macroblocks. By using macroblock groups and allowing some macroblocks to share prediction parameters while others have independent parameters, the system applies intra prediction partially where most beneficial, thereby improving overall prediction accuracy while reducing the total number of bits required compared to applying intra prediction to every single macroblock independently.
Data Source
AI summary
The present invention includes obtaining block type identification information on a partition of the current macroblock when the current macroblock is intra-mixed, specifying a bock type of the current macroblock based on the block type identification information, obtaining an intra prediction mode of the current macroblock according to the block type of the partition of the current macroblock, and predicting the current macroblock using the intra prediction mode and a pixel value of a neighboring block.