Transform Block Decoding via Single Truncated Unary Binarization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video compression standards, such as HEVC, face challenges in efficiently encoding and decoding transformed blocks of video samples, particularly in managing transform types for horizontal and vertical transforms, which affects compression performance and implementation costs in modern silicon processes.
Innovation Solution
A method and system for decoding a transform block in an image frame from a bitstream using a single truncated unary binarization to determine the type of horizontal and vertical transforms, allowing for efficient application of these transforms to the block, thereby reducing signaling overhead and improving compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple transforms are available for use in horizontal and vertical processing, then compression performance is improved, but signaling overhead increases due to the need to code indices to select which transform is to be used
Solution Approach 1:
The patent merges the selection of horizontal and vertical transforms into a single unified index. Instead of separately signaling horizontal transform index and vertical transform index, a single joint index is used that simultaneously identifies both transform types, thereby reducing signaling overhead while maintaining the ability to select from multiple transform combinations for optimal compression performance
Solution Approach 2:
The single joint index serves multiple functions: it identifies the horizontal transform type, identifies the vertical transform type, and implicitly defines the relationship between them. This multi-functional approach allows the system to maintain flexibility in transform selection while minimizing the number of bits required for signaling
2Loss of information
If a single truncated unary binarisation is used for both horizontal and vertical transforms, then signaling overhead is reduced, but the ability to precisely control individual transform selection is limited
Solution Approach 1:
The joint transform index is segmented into multiple fields or sub-indices that can be independently interpreted. The single truncated unary binarisation is divided into segments that respectively represent horizontal transform selection and vertical transform selection, allowing precise control over individual transform choices while maintaining the compact single-index structure for efficient signaling
Data Source
AI summary
The present disclosure provides a method 1200 of decoding a transform block in an image frame from a bitstream. The method comprises decoding a single truncated unary binarisation from the bitstream, the single truncated unary binarisation being used for a horizontal transform and a vertical transform of a transform block of the image frame. The method then determines a type of the horizontal and vertical transform based on the decoded single truncated unary binarisation and decodes the transform block in the image frame by applying the determined type of horizontal and vertical transform to the transform block of the image.