Continuous-Coordinate Motion Estimation for Subpixel Precision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion estimation and compensation methods using discrete coordinate systems require high memory and computational costs to capture subtle movements, limiting precision and efficiency in encoding and decoding multidimensional signals.
Innovation Solution
Employing a continuous coordinate system with fractional coordinates and on-the-fly resampling techniques to calculate motion vectors, allowing for precise motion compensation without the need for supersampled reference images, thus reducing memory and computational requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If discrete coordinate systems with integer coordinates are used for motion estimation, then device complexity is reduced, but measurement precision of motion vectors deteriorates
Solution Approach 1:
The patent changes the parameter of coordinate representation from discrete integer values to continuous fractional values. This allows motion vectors to specify positions with sub-pixel precision (e.g., 1/16th or 1/32nd of a pixel), directly improving measurement precision while maintaining compatibility with standard image processing architectures.
Solution Approach 2:
The patent introduces an additional dimension of precision by fractional coordinates between integer grid points. Instead of being constrained to discrete pixel locations, motion estimation operates in a continuous space, effectively adding a dimensional layer of precision without fundamentally changing the underlying image data structure.
2Measurement precision
If supersampled reference images are used to capture subtle movements, then measurement precision improves, but memory requirements increase
Solution Approach 1:
The patent performs preliminary resampling operations only on the specific regions of the reference image that are needed for motion estimation, rather than pre-processing the entire image at higher resolution. This selective approach captures subtle movements in areas of interest while avoiding the memory overhead of storing complete supersampled reference images.
Solution Approach 2:
The patent introduces fractional coordinate calculations as an intermediary mechanism between the discrete reference image and the motion estimation process. This intermediary layer enables sub-pixel precision without requiring the reference image itself to be stored at higher resolution, thus reducing memory requirements while maintaining measurement precision.
3Manufacturing precision
If high-resolution reference images are used for motion compensation, then manufacturing precision of motion estimation improves, but use of energy increases
Solution Approach 1:
The patent applies high-precision fractional coordinate calculations only locally to the specific blocks or regions undergoing motion estimation, rather than processing the entire image at full precision. This localized approach maintains manufacturing precision for motion compensation where needed while reducing overall computational energy consumption.
Solution Approach 2:
The patent uses fractional precision (excessive action) only when and where it is necessary for accurate motion estimation, rather than applying it uniformly across the entire image processing pipeline. This partial application of high-precision calculations reduces energy consumption while maintaining sufficient precision for capturing subtle movements.
4Measurement precision
If continuous coordinate systems with fractional coordinates are employed, then measurement precision improves, but device complexity increases
Solution Approach 1:
The patent replaces complex hardware-based continuous coordinate processing with software-based fractional arithmetic operations. By implementing fractional coordinate calculations through standard computational algorithms rather than specialized hardware, the system achieves high measurement precision while avoiding a proportional increase in device complexity.
Solution Approach 2:
The patent designs the fractional coordinate system to be universally applicable across different motion estimation algorithms and image processing architectures. This multi-functional approach allows the same fractional coordinate mechanism to serve multiple purposes (motion estimation, motion compensation, interpolation) without requiring separate complex processing paths for each function.
Data Source
AI summary
Computer processor hardware receives settings information for a first image. The first image includes a set of multiple display elements. The computer processor hardware receives motion compensation information for a given display element in a second image to be created based at least in part on the first image. The motion compensation information indicates a coordinate location within a particular display element in the first image to which the given display element pertains. The computer processor hardware utilizes the coordinate location as a basis from which to select a grouping of multiple display elements in the first image. The computer processor hardware then generates a setting for the given display element in the second image based on settings of the multiple display elements in the grouping.