Motion Vector Estimation for Frame Rate Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-speed frame rate conversion in image processing often results in broken images, edge shaking, or judder due to the imbalance between calculation cost and accuracy in motion vector estimation.
Innovation Solution
The method involves downsizing frames into blocks, performing first and second motion estimations using 3D recursive searches, and adjusting estimation parameters based on initial motion vectors to generate accurate motion vectors efficiently, thereby reducing computational cost and image artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If high-speed frame rate conversion is performed with limited calculation cost, then processing speed is improved, but image quality deteriorates due to broken images, edge shaking, or judder
Solution Approach 1:
The patent divides the frame rate conversion process into multiple stages: a first motion estimation stage that processes all blocks, and a second motion estimation stage that selectively processes only candidate blocks. This segmentation allows the system to maintain high processing speed while improving image quality for critical regions by applying more rigorous estimation only where needed.
Solution Approach 2:
The patent performs a preliminary first motion estimation on all blocks to generate initial motion vectors before conducting the second motion estimation. This preliminary action provides baseline motion information that guides the subsequent selective refinement process, enabling the system to identify candidate blocks that require more accurate estimation without processing all blocks at full detail.
2Manufacturing precision
If motion estimation accuracy is improved through comprehensive processing, then image quality is improved, but calculation cost increases
Solution Approach 1:
The patent applies different processing quality levels to different blocks based on their characteristics. Candidate blocks identified through the first motion estimation undergo a second, more accurate motion estimation, while non-candidate blocks use only the first estimation results. This local quality approach ensures high accuracy for critical blocks while maintaining low calculation cost for the overall frame.
Solution Approach 2:
Instead of performing comprehensive motion estimation on all blocks (excessive action), the patent applies the second motion estimation only to selected candidate blocks (partial action). This partial processing approach achieves sufficient accuracy for the most critical regions while avoiding the excessive computational cost of processing every block at maximum detail.
3Device complexity
If selective motion estimation is performed on candidate blocks only, then calculation cost is reduced, but motion vector accuracy may deteriorate
Solution Approach 1:
The patent uses feedback from the first motion estimation results to identify candidate blocks that require second motion estimation. The initial motion vectors serve as feedback that guides the selection process, ensuring that blocks with potentially inaccurate or uncertain motion information are selected for refined estimation, thereby maintaining overall accuracy while reducing calculation cost.
Solution Approach 2:
The first motion estimation serves as a preliminary screening process that identifies which blocks need further processing. By establishing this preliminary assessment, the system can confidently apply selective second motion estimation to only those blocks where it is most beneficial, maintaining accuracy for critical regions while avoiding unnecessary processing elsewhere.
Data Source
AI summary
An image processing method includes: downsizing a current frame and a reference frame; dividing the down-sized current frame and the down-sized reference frame into multiple first current blocks and multiple first reference blocks, respectively; performing a first motion estimation to the first current blocks and the first reference blocks to generate multiple first motion vectors; dividing the current picture and the reference picture into multiple second current blocks and multiple second reference blocks, respectively; performing a second motion estimation to the second current blocks and the second reference blocks to generate multiple second motion vectors; and generating a compensated frame between the current frame and the reference frame according to the second motion vectors. The second motion estimation includes: performing a 3D recursive search for each second current block; and adjusting multiple estimation parameters in the 3D recursive search according to the first motion vector.