Motion Vector Rounding for Video Coding Precision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding methods, particularly in direct mode, face challenges in accurately determining motion vectors due to decimal fractions, leading to reduced compression efficiency and image prediction accuracy, especially in B frame coding where 70%-80% of modes are occupied by direct mode.
Innovation Solution
A method is introduced to calculate forward and backward motion vectors using specific formulas that involve a scale_factor and shift_len to maintain exact motion vector values without division, improving precision and coding efficiency by using rounding to zero, and utilizing these vectors to determine reference image blocks for more accurate prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If rounding method is used to convert decimal motion vectors to integers, then motion vectors can be processed, but prediction accuracy deteriorates and bias appears in image blocks
Solution Approach 1:
The patent introduces a new dimension of sub-pixel precision by allowing motion vectors to operate at 1/4 pixel resolution instead of being constrained to integer pixel positions. This dimensional extension enables more precise reference block selection without requiring complex rounding operations, thereby maintaining both ease of processing and high prediction accuracy.
Solution Approach 2:
The patent changes the precision parameter of motion vectors from integer pixels to sub-pixel values (1/4 pixel precision). By modifying this fundamental parameter, the system achieves better prediction accuracy while maintaining computational feasibility through standardized sub-pixel interpolation methods.
2Productivity
If direct mode is used to reduce bit usage for motion vectors, then coding efficiency improves, but motion vector deduction accuracy deteriorates
Solution Approach 1:
The patent enhances the precision parameter of deduced motion vectors by implementing sub-pixel calculation mechanisms. Through improved interpolation algorithms and extended precision arithmetic, the system achieves higher accuracy in motion vector deduction while maintaining the bit-efficient direct mode structure.
3Device complexity
If integer motion vectors are used, then computation is simpler, but prediction accuracy deteriorates due to inability to represent decimal positions
Solution Approach 1:
The patent extends the position representation from integer pixel coordinates to sub-pixel coordinates (1/4 pixel precision). This dimensional extension allows motion vectors to point to more precise locations in the reference frame, improving prediction accuracy while maintaining computational tractability through standardized sub-pixel interpolation.
Solution Approach 2:
The patent introduces sub-pixel interpolation as an intermediary mechanism that bridges integer motion vector indices and continuous pixel positions. This intermediary layer enables precise reference block extraction without requiring complex floating-point computations, thus maintaining computational simplicity while achieving high positioning accuracy.
Data Source
AI summary
A “rounding to zero” method can maintain the exact motion vector and can also be achieved by the method without division so as to improve the precision of calculating the motion vector, embody the motion of the object in video more factually, and obtain the more accurate motion vector prediction. Combining with the forward prediction coding and the backward prediction coding, the present invention realizes a new prediction coding mode, which can guarantee the high efficiency of coding in direct mode as well as is convenient for hardware realization, and gains the same effect as the conventional B frame coding.