Temporal Motion Vector Scaling for Video Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The motion vector used in the temporal direct mode in existing moving picture coding schemes is limited to a forward direction in display time order, making it difficult to derive an appropriate motion vector, which decreases the compression rate.
Innovation Solution
A moving picture coding method that determines whether a co-located block has two reference motion vectors, calculates motion vector predictor candidates by scaling these vectors, and selects the predictor with the smallest difference to code the current block, along with the difference value and predictor index, to enhance compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If the motion vector is limited to forward direction in display time order, then the coding process is simple, but the compression rate decreases
Solution Approach 1:
The patent makes the motion vector direction dynamic by allowing selection between forward and backward reference pictures based on actual motion characteristics. The system determines the appropriate reference picture direction (forward or backward) and uses corresponding motion vectors accordingly, rather than being fixed to one direction. This dynamic adaptation enables better compression while maintaining manageable complexity through structured decision-making.
Solution Approach 2:
The patent changes the parameter of motion vector direction from fixed (forward only) to variable (forward or backward). By allowing the motion vector to be selected based on the actual motion characteristics of the current picture, the system optimizes compression efficiency. The reference picture direction is changed as a variable parameter that can be adjusted according to the specific coding situation.
2Device complexity
If only one reference picture list is used, then the device complexity is low, but the adaptability to different motion directions is limited
Solution Approach 1:
The patent segments the reference picture list into multiple lists (first reference picture list and second reference picture list), where each list corresponds to a specific reference picture direction (forward or backward). This segmentation allows the system to store and manage reference pictures organized by direction, enabling selective access to appropriate reference pictures based on the current motion characteristics while maintaining a structured and manageable organization.
Solution Approach 2:
The patent creates a multi-functional reference picture list structure where both forward and backward reference pictures can be stored and accessed. The system can selectively use either the first or second reference picture list depending on the motion direction, making the overall system adaptable to various motion scenarios. This universal structure allows a single coding system to handle multiple motion directions effectively.
Data Source
AI summary
A moving picture coding method includes: determining whether or not a first block has two reference motion vectors for referencing a same direction that is one of a forward direction and a backward direction in display order, the first block being included in a first picture different from the current picture and being co-located, in the first picture, with the current block within the current picture; and calculating, when it is determined in the determining that the first block has the two reference motion vectors, first and second motion vector predictor candidates that are candidates for a motion vector for the current block, by scaling the two reference motion vectors.