Video Filter Shapes With Fewer Taps for Sample Offset Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face inefficiencies in representing intra prediction directions, particularly those that are statistically less likely to occur in video content, leading to increased bit usage and reduced compression efficiency.
Innovation Solution
Implementing a non-linear mapping based filter with a configurable filter shape configuration, using offset values to apply filters on samples, including cross-component and local sample offset filters, to enhance compression efficiency by optimizing the representation of less likely intra prediction directions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional linear mapping is used for intra prediction directions, then the coding structure is simple, but the bit usage increases for less likely directions and compression efficiency decreases
Solution Approach 1:
The patent applies non-linear mapping to transform the relationship between intra prediction direction indices and actual prediction angles. By changing the mapping parameters from linear to non-linear, the system optimizes bit allocation such that less likely prediction directions use fewer bits, directly improving compression efficiency while reducing overall bit usage for representation
Solution Approach 2:
The patent introduces dynamic adaptation by allowing the non-linear mapping characteristics to be adjusted based on content statistics. The mapping function can be modified to reflect the actual distribution of prediction directions in the video content, making the bit allocation dynamic rather than static, thereby optimizing compression for different content types
2Measurement precision
If more filter taps are used in the non-linear mapping based filter, then the filtering precision improves, but the computational complexity increases
Solution Approach 1:
The patent implements a configurable filter shape configuration that allows selective application of different numbers of filter taps based on the specific filtering needs. Rather than always applying maximum filtering (excessive action), the system applies only the necessary degree of filtering (partial action) to achieve the required precision while minimizing computational complexity
Solution Approach 2:
The patent divides the filtering operation into multiple configurable stages with different filter shapes and tap configurations. This segmentation allows the system to apply different levels of filtering precision to different regions or prediction directions, optimizing the balance between precision and complexity by not uniformly applying high complexity filtering everywhere
Data Source
AI summary
In a method for filtering in video decoding, a coded video bitstream that includes filter shape index information is received. A filter shape configuration of a video filter is determined based on the filter shape index information. The filter shape index information indicates that a number of filter taps of the filter shape configuration is less than five. Offset values associated with the filter shape configuration of the video filter are determined. The video filter is applied on a sample based on the offset values associated with the filter shape configuration.