Adaptive Filter Selection for Scalable Video Coding Phase Mismatch
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding systems lack flexibility in up-sampling and down-sampling processes, leading to coding efficiency losses due to incorrect phase selection during inter-layer prediction in scalable video coding, particularly in HEVC, where a single fixed up-sampling filter is used regardless of phase mismatch between layers.
Innovation Solution
The system determines phase displacement information between video layers and uses adaptive image filter sets based on this information to perform up-sampling and down-sampling, embedding phase information in the bit stream to select appropriate filters and eliminate coding efficiency losses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single fixed up-sampling filter is used in HEVC inter-layer prediction, then device complexity is reduced, but coding efficiency deteriorates due to phase mismatch between layers
Solution Approach 1:
The patent applies dynamics by transitioning from a static fixed filter approach to a dynamic adaptive filter selection mechanism. The system determines phase displacement information between base layer and enhancement layer, then adaptively selects from multiple up-sampling filter sets (first set for zero-phase, second set for symmetric-phase) based on the determined phase relationship. This dynamic adaptation resolves the contradiction by allowing the system to maintain low complexity through predefined filter sets while achieving high coding efficiency through phase-aware filter selection.
Solution Approach 2:
The patent changes the parameter of filter selection from fixed to variable based on phase displacement information. By introducing phase displacement determination and using this parameter to select between different filter sets, the system optimizes up-sampling performance for different phase conditions without significantly increasing overall system complexity, as the filter sets themselves are predetermined.
2Measurement precision
If phase displacement information is signaled in the bit stream, then inter-layer prediction accuracy is improved, but bit stream complexity increases
Solution Approach 1:
The patent applies preliminary action by determining phase displacement information in advance during the encoding process and signaling it in the bit stream before actual inter-layer prediction occurs. This allows the decoder to pre-configure the appropriate up-sampling filter set based on the signaled phase information, ensuring accurate phase alignment without adding complex real-time decision-making structures to the bit stream.
Solution Approach 2:
The phase displacement information acts as an intermediary parameter that bridges the base layer and enhancement layer in inter-layer prediction. By signaling this intermediate information in the bit stream, the system enables the decoder to select the appropriate filter set, thereby improving prediction accuracy while keeping the bit stream structure relatively simple through the use of efficient syntax elements.
3Productivity
If adaptive image filter sets are used based on phase displacement information, then up-sampling performance is improved, but processing time increases
Solution Approach 1:
The patent applies preliminary action by pre-defining multiple up-sampling filter sets (first up-sampling filter set for zero-phase displacement, second up-sampling filter set for symmetric-phase displacement) before the actual encoding/decoding process. The phase displacement information is determined and used to select the appropriate pre-defined filter set, avoiding the need for complex real-time filter design or optimization, thus improving up-sampling performance while limiting processing time overhead.
Solution Approach 2:
The patent changes the parameter of filter selection from a single fixed filter to multiple adaptive filter sets based on phase displacement parameters. By organizing filters into predetermined sets corresponding to different phase conditions, the system achieves better up-sampling performance through parameter-adaptive selection while controlling processing time through efficient parameter-based indexing rather than complex optimization routines.
Data Source
AI summary
Methods and systems for video image coding are provided. Sets of filters may be selected and applied to video information at least partially based on phase displacement information between a first and second layer of video information. For example, the phase displacement information may correspond to a difference between a position of a pixel in the first layer and a corresponding position of the pixel in the second layer. The selected filter set can be an up-sampling filter or a down-sampling filter. The phase displacement information may be encoded as a syntax element embedded in the video bit stream.