Image Coding Merge Candidates with Predetermined Zero Motion Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image coding methods using inter prediction lack sufficient coding efficiency, particularly in handling stationary regions and varying motion vectors.
Innovation Solution
An image coding method that derives merging candidates based on neighboring blocks and includes predetermined motion vectors, such as zero vectors, to enhance coding efficiency by increasing the variety of candidates and allowing parallel derivation and decoding of indices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional inter prediction coding is used with motion estimation, then motion vectors can be derived for moving regions, but coding efficiency is insufficient for stationary regions and the complexity of motion estimation increases processing time
Solution Approach 1:
The patent segments the candidate list into two distinct parts: first candidates derived from spatial and temporal neighboring blocks, and second candidates with predetermined motion vectors (including zero vectors for stationary regions). This segmentation allows the system to handle moving and stationary regions differently, improving coding efficiency without requiring complex motion estimation for all blocks.
Solution Approach 2:
The patent performs preliminary action by pre-defining a set of candidate motion vectors (including zero vectors, identity matrices, and other predetermined values) before the actual coding process. These predetermined candidates are prepared in advance and directly available for selection, eliminating the need for complex motion estimation in stationary regions and reducing processing time.
2Productivity
If more merging candidates are derived from neighboring blocks, then the variety of candidates increases and coding efficiency improves, but the amount of information to be coded increases
Solution Approach 1:
The patent applies local quality by assigning different characteristics to different parts of the candidate list. First candidates (from neighboring blocks) are suitable for regions with similar motion characteristics, while second candidates (predetermined vectors) are optimized for stationary regions or regions with specific motion patterns. This localized optimization improves coding efficiency without requiring all candidates to be fully coded.
Solution Approach 2:
The patent changes parameters by introducing a flag-based selection mechanism. Instead of coding all candidate parameters, the system uses flags to indicate which candidate is selected, significantly reducing the amount of information to be coded while maintaining the ability to represent a wide variety of motion scenarios through the diverse candidate list.
3Measurement precision
If motion estimation is performed on all blocks, then accurate motion vectors are obtained, but the processing time and computational load increase significantly
Solution Approach 1:
The patent applies partial action by performing motion estimation only when necessary (for first candidates from neighboring blocks) rather than on all blocks. The predetermined second candidates provide sufficient accuracy for many cases (especially stationary regions) without requiring full motion estimation, reducing processing time while maintaining adequate accuracy through the diverse candidate selection.
Data Source
AI summary
A method for transmitting a bitstream via a network is provided. The bitstream being generated by: deriving a first candidate having a first motion vector that has been used to code a first block; deriving a second candidate having a second motion vector and a first reference picture index value that identifies a first reference picture corresponding to the second motion vector, with the second motion vector being a first zero vector; deriving a third candidate having a third motion vector and a second reference picture index value that identifies a second reference picture corresponding to the third motion vector, with the third motion vector being a second zero vector. One candidate from a plurality of candidates, including the first, second, and third candidate is selected, and an index identifying the selected one candidate is coded.