Sub-pixel Interpolation Using Segmented Filter Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video encoding techniques face challenges in accurately predicting motion vectors at sub-pixel granularity, requiring additional pixel values that are resource-intensive to generate, especially in hardware-constrained environments.
Innovation Solution
The solution involves using lower complexity filters to compute half-pixels and diagonal half-pixels, followed by higher complexity filters for sub-pixel motion estimation, reducing computational requirements while maintaining accuracy, particularly for quarter and eighth pixel resolutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If higher complexity filters are used to generate sub-pixel values, then measurement precision of motion vectors is improved, but device complexity and resource requirements increase
Solution Approach 1:
The patent segments the sub-pixel generation process into multiple stages: first generating half-pixel values using lower complexity filters, then generating quarter-pixel values using those half-pixel values. This hierarchical segmentation allows achieving high precision (quarter-pixel level) while using simpler filters at each stage, thereby reducing overall device complexity while maintaining measurement precision.
Solution Approach 2:
The patent performs preliminary generation of half-pixel values using lower complexity filters before generating the final quarter-pixel sub-pixel values. This preliminary action creates intermediate results that facilitate the final high-precision motion vector calculation without requiring the most complex filters to be applied directly to all pixel positions, thus reducing device complexity while maintaining precision.
2Manufacturing precision
If higher complexity filters are used to generate sub-pixel values, then manufacturing precision of interpolated pixels is improved, but resource requirements increase
Solution Approach 1:
The patent divides the interpolation process into segmented stages where half-pixel values are computed first using fewer computational resources, and then quarter-pixel values are derived from those half-pixel values. This segmentation reduces the total computational resources required while maintaining manufacturing precision of the final interpolated pixels.
Solution Approach 2:
The patent applies lower complexity filters partially - specifically using them for generating half-pixel values which are then used to compute quarter-pixel values. This partial application of filter complexity achieves sufficient interpolated pixel accuracy without requiring full complexity filters to be applied throughout the entire process, thereby reducing resource requirements.
3Measurement precision
If sub-pixel motion estimation is performed, then motion vector accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent implements segmented sub-pixel motion estimation where quarter-pixel motion vectors are derived from half-pixel values which themselves are derived from integer pixel values. This multi-stage segmentation enables accurate sub-pixel motion estimation while keeping computational complexity manageable by using simpler operations at each stage rather than applying complex operations all at once.
Solution Approach 2:
The patent performs preliminary computation of half-pixel values using lower complexity filters before performing the final quarter-pixel motion estimation. This preliminary action reduces the computational burden of the main motion estimation task while still achieving high accuracy, thereby improving motion vector accuracy without proportionally increasing overall computational complexity.
4Measurement precision
If sub-pixel motion estimation is performed, then motion vector accuracy is improved, but resource requirements increase
Solution Approach 1:
The patent segments the sub-pixel motion estimation into hierarchical stages (integer pixels → half-pixels → quarter-pixels), where each stage uses the results of the previous stage. This segmentation reduces the total computational resources required compared to directly computing quarter-pixel values from integer pixels, while still achieving high motion vector accuracy.
Solution Approach 2:
The patent uses lower complexity filters partially in the sub-pixel motion estimation process - specifically for generating half-pixel values that are then used for quarter-pixel motion vectors. This partial use of filter complexity achieves sufficient motion vector accuracy without requiring the most computationally intensive operations throughout the entire process, thereby reducing resource requirements.
Data Source
AI summary
Reducing computational complexity when generating sub-pixel values for sub-pixel motion estimation from integer pixels. In an embodiment, half pixels in vertical and horizontal directions are computed by a applying a filter of first complexity on integer pixels, and a half pixel in diagonal direction is computed using a filter of lower complexity as compared to the filter of first complexity. Quarter (and other lower resolution pixels) pixels may also be generated using the half pixel in the diagonal direction. Thus, overall computational complexity is reduced in generating sub-pixels for sub-pixel motion estimation.