Video Decoding Candidate List Selection by Reference Mode Difference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video coding methods face challenges in quickly selecting an appropriate prediction mode for block units due to a high number of intra modes, leading to differences in frequent mode lists for adjacent and separated neighboring blocks, which are not adapted to varying coding blocks effectively.
Innovation Solution
A method for decoding a bitstream that determines a block unit by selecting a prediction mode from candidate lists based on the difference between reference mode indices, where the difference is used to choose between specific candidate lists, ensuring accurate reconstruction of block units by adapting to different coding blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a high number of intra modes are used for prediction, then prediction accuracy is improved, but the time required to select an appropriate prediction mode increases
Solution Approach 1:
The patent divides the large set of intra modes into multiple candidate lists (first candidate list, second candidate list, third candidate list) with different numbers of modes. Different lists are used based on the specific coding block characteristics, allowing the system to segment the search space and avoid evaluating all modes for every block, thus reducing selection time while maintaining accuracy where needed.
Solution Approach 2:
The patent dynamically adjusts the candidate list used for prediction based on the difference between reference mode indices of neighboring blocks. When the difference is small (indicating similar neighboring modes), a smaller candidate list is used for faster selection. When the difference is large, a larger candidate list is used to ensure prediction accuracy. This dynamic adaptation resolves the contradiction between speed and accuracy.
2Device complexity
If a fixed frequent mode list is used for all coding blocks, then the encoding process is simplified, but the adaptation to different coding blocks is reduced
Solution Approach 1:
The patent implements a dynamic candidate list selection mechanism that adapts to different coding blocks based on reference mode index differences. Instead of using a single fixed list for all blocks, the system selects from multiple predefined lists (with 6, 12, or 18 modes) depending on the specific block's characteristics, thereby maintaining both simplicity and adaptability.
Solution Approach 2:
The patent applies different candidate lists to different coding blocks based on their local characteristics (reference mode index differences). Each block receives a tailored candidate list appropriate to its specific context rather than a uniform approach, improving adaptation while maintaining overall system simplicity through the use of predefined lists.
3Measurement precision
If different candidate lists are selected based on reference mode index difference, then prediction accuracy is improved, but the complexity of determining which list to use increases
Solution Approach 1:
The patent uses a simple parameter (difference between reference mode indices) to trigger selection from predefined candidate lists. The complexity is managed by establishing clear thresholds (difference ≤ 1, difference ≤ 2, difference ≤ 3) that map to specific lists, transforming a potentially complex selection problem into a straightforward parameter-based decision process.
Data Source
AI summary
A method of decoding a bitstream by an electronic device is provided. An image frame is received from the bitstream, and a block unit is determined from the image frame. A prediction index of the block unit is determined from the bitstream. A plurality of reference mode indices is determined based on a plurality of neighboring blocks of the block unit. The electronic device determines whether the reference mode indices are greater than a specific one of a plurality of non-angular mode indices. When the reference mode indices are greater than the specific non-angular mode index, the electronic device determines whether a difference between the reference mode indices is equal to two. When the difference is equal to two, a prediction mode is selected according to the prediction index and based on a first candidate list. Then, the block unit is reconstructed based on the prediction mode.