Motion Vector Estimation Pipeline for Low Bit Rate Video
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion vector estimation methods face challenges in maintaining image quality at low bit rates and are difficult to implement in a pipeline structure due to the need for pre-determined past motion vectors for calculating addition amounts.
Innovation Solution
A motion vector estimating device that segments input pictures into blocks and uses an n-stage pipeline processing to calculate motion vectors based on distortion amounts and correction values, allowing for a trade-off between bit amounts and differential amounts, and enabling high-speed estimation with a pipeline structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If the bit amount related to motion vector is reduced at low bit rate, then the bit rate is reduced, but the image quality drastically deteriorates
Solution Approach 1:
The patent changes the parameter of motion vector representation by using differential encoding with variable bit allocation. Instead of fixed bit representation, the system dynamically adjusts the number of bits used for motion vector differential amounts based on the balance between bit rate constraints and image quality requirements, allowing optimization at low bit rates while maintaining acceptable quality.
Solution Approach 2:
The patent implements feedback through the use of previously estimated motion vectors to predict current motion vectors. The differential amount is calculated based on the difference between current and previous motion vectors, creating a feedback loop that adapts the motion vector representation to historical data, thereby reducing the bits needed while maintaining quality through intelligent prediction.
2Productivity
If conventional motion vector estimation method is used, then motion vector can be estimated, but it is difficult to employ pipeline structure because past motion vectors must be determined in advance
Solution Approach 1:
The patent applies preliminary action by estimating motion vectors in advance for subsequent blocks. The motion vector for a block is estimated based on the previously determined motion vector of the left neighboring block, allowing the pipeline to proceed without waiting for all motion vectors to be fully calculated, thus enabling high-speed estimation while maintaining pipeline structure.
Solution Approach 2:
The patent segments the motion vector estimation process into independent units that can be processed in parallel. Each block's motion vector is estimated independently based on its neighboring blocks, allowing the pipeline to process multiple blocks simultaneously at different stages, thereby achieving high productivity without excessive complexity.
3Ease of operation
If addition amount is calculated from motion vector variation, then motion vector can be coded, but the addition amount must be determined in advance which prevents pipeline structure
Solution Approach 1:
The patent makes the addition amount calculation dynamic by using the actual previously determined motion vector values rather than fixed pre-calculated amounts. The addition amount is calculated as α·(|vnx−pvx|+|vny—pvy|) where pvx and pvy are the actual previous motion vector components, allowing flexible adaptation during pipeline processing without requiring advance determination, thus enabling both ease of operation and high productivity.
Data Source
AI summary
To provide a motion vector estimating device which is capable of calculating the optimum motion vector by a trade-off between the bit rate of a motion vector and a differential amount, and estimating a motion vector in a pipeline structure at high speed, a motion vector estimating method which enables the same. The motion vector estimating device (100) includes: a first motion vector estimating unit (102) which calculates an evaluation value based on the amount of distortion calculated using search blocks and an input block and a correction value calculated using a motion vector with respect to each search block, a first prediction vector, and a weighting coefficient, and estimates a first motion vector of the input block based on the location of the input block and the location of a search block which provides the smallest evaluation value; a weighted coefficient calculating unit (103) which calculates the weighted coefficient; a prediction vector generating unit (104) which generates the first motion vector based on the first motion vector; and a second motion estimating unit (105) which estimates a motion vector determining, as an evaluation value, only the amount of distortion calculated using the search blocks and the input block.