Motion Vector Selection by Temporal Direction for Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion vector calculation methods in temporal direct are limited by using only forward-directed motion vectors from reference pictures after the current block, making it difficult to determine the most suitable motion vector, leading to decreased compression rates in video coding.
Innovation Solution
A method to select and calculate motion vectors based on the position and temporal distance of reference blocks relative to the current block, allowing selection of forward or backward vectors as needed, and scaling them to derive the most suitable motion vector for the current block.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If only forward-directed motion vectors from reference pictures after the current block are used in temporal direct, then the motion vector calculation is simplified, but the compression rate decreases
Solution Approach 1:
The patent makes the motion vector selection dynamic by introducing a flag that indicates whether to use forward or backward reference pictures. This allows the system to adaptively choose the most suitable reference picture direction based on the specific coding situation, rather than being fixed to only forward-directed vectors, thereby improving compression rate while maintaining calculation simplicity.
Solution Approach 2:
The patent changes the parameter of motion vector direction by introducing a flag that can take different values to select between forward and backward reference pictures. This parameter change enables the system to switch between different motion vector calculation modes, resolving the contradiction between simplicity and compression efficiency.
2Measurement precision
If motion vectors are selected based on temporal distance and position of reference blocks, then the most suitable motion vector can be determined, but the selection process becomes more complex
Solution Approach 1:
The patent applies local quality by considering the specific position and temporal distance of reference blocks relative to the current block. Instead of using a uniform selection criterion, the system evaluates the local characteristics of each reference block and selects motion vectors based on these local properties, thereby improving accuracy without requiring complex global optimization.
Solution Approach 2:
The patent performs preliminary action by pre-calculating and storing temporal distance information and position flags for reference blocks. This preliminary preparation allows the motion vector selection to be made quickly during coding without requiring complex real-time calculations, thus improving accuracy while keeping the selection process manageable.
3Productivity
If scaling is applied to reference motion vectors to derive the most suitable motion vector, then coding efficiency improves, but calculation operations increase
Solution Approach 1:
The patent applies partial action by using scaling only when necessary to derive the most suitable motion vector from reference motion vectors. The scaling factor is calculated based on temporal distances, and scaling is applied selectively rather than universally, thereby improving coding efficiency while minimizing the increase in calculation operations.
Data Source
AI summary
A motion vector calculation method which attains a higher compression rate, includes: a selection step of selecting one of at least one reference motion vector of a reference block; and a calculation step of calculating a motion vector of a current block to be processed, using the one reference motion vector selected in the selection step, and in the selection step, when the reference block has two reference motion vectors, one of the two reference motion vectors is selected based on whether the reference block is located before or after the current block in display time order, and when the reference block has only one reference motion vector, the one reference motion vector is selected.