Fixed Bit Depth Processing for CCLM Video Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing video encoding and decoding techniques, particularly those using cross-component linear modeling (CCLM) mode, require significant hardware resources due to their calculation intensity and the need for large value storage, especially for higher bit depths.
Innovation Solution
By decoding blocks in the CCLM mode using a fixed bit depth different from the block's bit depth and downsampling the luma block to determine intermediate luma blocks for deriving CCLM parameters, the hardware resources required for CCLM mode can be reduced.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If CCLM mode is implemented with higher bit depths, then prediction accuracy is improved, but hardware resources and computational complexity increase significantly
Solution Approach 1:
The patent applies parameter changes by using a fixed bit depth (10 bits) for CCLM parameter derivation regardless of the actual input bit depth. This transforms the variable parameter (bit depth) into a fixed value, thereby reducing computational complexity and hardware resource requirements while maintaining acceptable prediction accuracy through the fixed-parameter approach
2Reliability
If CCLM mode processes data at original bit depth, then prediction quality is maintained, but hardware resource consumption increases
Solution Approach 1:
The patent creates a simplified intermediate representation at fixed bit depth (10 bits) that serves as a disposable working model for parameter derivation. This intermediate low-precision representation is used temporarily to compute CCLM parameters, after which the original high-precision data is processed using these pre-computed parameters, thereby reducing hardware resource consumption while maintaining prediction quality
Data Source
AI summary
A video decoder can be configured to determine that a first block of the video data is encoded in a cross component linear model (CCLM) mode; determine that the first block has a first bit depth; determine that a second block of the video data is encoded in the CCLM mode; determine that the second block has a second bit depth that is different than the first bit depth; and decode the first block and the second block in the CCLM mode using a fixed bit depth, wherein the fixed bit depth is different than at least one of the first bit depth or the second bit depth.