Adaptive Motion Estimation for Video Format Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion estimation technologies face challenges in extracting reliable motion vectors, particularly in video format conversion, where they struggle to maintain accuracy and precision, leading to suboptimal picture quality and inefficiencies in spatiotemporal interpolation.
Innovation Solution
A motion estimation apparatus and method that calculates block errors using candidate motion vectors from previous and adjacent frames, employing adaptive motion estimation modes based on variance, average compensation error, and minimal Sum of Absolute Difference (SAD) to estimate final motion vectors, incorporating half-pixel, three-stage search, and block split modes for enhanced reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional block matching algorithm is used for motion estimation, then motion vector can be estimated, but reliability of motion vector is insufficient in video format conversion
Solution Approach 1:
The patent implements dynamic motion estimation mode selection that adapts to different block characteristics. The system calculates block errors and variance for each block, then dynamically chooses between quarter-pixel, half-pixel, or integer-pixel precision modes based on the calculated metrics, ensuring optimal reliability for each block's specific motion characteristics
Solution Approach 2:
The patent changes the precision parameter of motion vector estimation based on block characteristics. By calculating block error and variance, the system adjusts the search precision dynamically - using quarter-pixel for high-precision needs, half-pixel for moderate needs, and integer-pixel for low-precision needs, thereby optimizing both reliability and computational efficiency
2Measurement precision
If high precision motion estimation is performed for all blocks, then motion vector accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies different motion estimation precision levels to different blocks based on their local characteristics. Blocks with high variance or large block errors receive quarter-pixel precision treatment, while blocks with low variance use integer-pixel precision. This local adaptation ensures high accuracy where needed while reducing computational complexity in regions with simpler motion patterns
Solution Approach 2:
The patent segments the video frame into multiple blocks and processes each block independently with appropriate precision levels. By dividing the frame into 16x16 macroblocks and further into 8x8 sub-blocks when necessary, the system can apply different computational strategies to different segments, optimizing the overall computational efficiency while maintaining necessary accuracy
3Productivity
If conventional motion estimation is used, then processing speed is maintained, but picture quality in video format conversion deteriorates
Solution Approach 1:
The patent performs preliminary calculations of block errors and variance before final motion vector estimation. By pre-calculating these metrics for each block, the system can quickly determine the appropriate precision level needed, avoiding unnecessary high-precision computations and maintaining processing speed while ensuring adequate picture quality through adaptive precision selection
Data Source
AI summary
A motion estimation apparatus and method is disclosed. The apparatus includes a block error calculator to calculate a specified block error based on a candidate motion vector previously estimated and spatiotemporarily neighboring on a present block; a motion estimation mode decider to decide a motion estimation mode for the present block to be processed based on the calculated block error; and a motion vector estimator to estimate a last motion vector in the decided motion estimation mode.