Image Interpolation Decision Mixing for Artifact Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Standard interpolation methods in digital image processing, particularly in video sequences, suffer from abrupt changes and artifacts due to competing candidate directions with similar loss values, leading to discontinuities and structure perception issues, especially in corner cases and fade-in/fade-out effects.
Innovation Solution
An image interpolation method that determines loss values for candidate interpolators, selects the best interpolator(s) by minimizing loss, and computes output pixel values using a weighted sum of interpolated values, avoiding abrupt switches and artifacts by using a smooth decision map and invariant weights with respect to contrast changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If standard interpolation methods select the direction with the smallest loss value, then the interpolation speed is maintained, but abrupt switches and discontinuities appear in the output images causing artifacts
Solution Approach 1:
The patent changes the decision parameter from a discrete selection (one direction with minimum loss) to a continuous parameter (mixing weights between multiple directions). By using a smooth transition function based on loss values to determine mixing proportions, the method maintains interpolation speed while eliminating abrupt switches that cause artifacts.
Solution Approach 2:
The patent combines multiple candidate directional interpolators into a composite interpolation result using weighted mixing. Instead of selecting a single direction, it creates a composite output by blending results from multiple directions according to their relative loss values, thereby maintaining image continuity while preserving computational efficiency.
2Adaptability or versatility
If candidate directions with similar loss values compete, then the adaptability to different image structures is improved, but discontinuities and perceived artifacts appear in the output
Solution Approach 1:
The patent transforms the discrete decision-making process into a continuous parameter optimization problem. By using a smooth transition function that continuously adjusts mixing weights based on loss value differences, it maintains adaptability to various image structures while ensuring output consistency through gradual transitions rather than abrupt switches.
Solution Approach 2:
The patent introduces a mixing weight as an intermediary parameter between candidate directions. This intermediary smoothly transitions the contribution of different directional interpolators based on their loss values, preventing direct abrupt switches between competing directions and thereby eliminating artifacts while preserving adaptability.
3Adaptability or versatility
If the interpolation processor changes decisions along fade-in or fade-out time, then the interpolation adapts to changing contrast, but the shape of the interpolated image changes strangely appearing as artifacts
Solution Approach 1:
The patent uses a smooth transition function that continuously adjusts the mixing weights based on changing loss values during fade-in/fade-out effects. This continuous parameter adjustment allows the interpolation to adapt to contrast changes while maintaining image shape stability through gradual transitions rather than abrupt decision changes.
Solution Approach 2:
The patent ensures continuous adjustment of mixing weights throughout the fade-in/fade-out process. By maintaining continuous useful action in the form of smooth weight transitions rather than discrete decision switches, the method adapts to contrast changes while preserving image shape consistency and eliminating strange artifacts.
Data Source
AI summary
The method comprises for a pixel of an output grid which is not on an input grid on which pixel values are received in an input signal: determining a loss value for each of a plurality of candidate interpolators by analyzing the pixel values of the input grid along a direction associated with the candidate interpolator; selecting at least one interpolator by minimizing the loss value; and determining an output pixel value. At least two interpolation modes are provided, including: a first mode in which one interpolator is selected and the output pixel value is determined as an interpolated value obtained by applying locally the selected interpolator to pixel values of the input grid; and a second mode in which more than one interpolator is selected and the output pixel value is determined as a weighted sum of a plurality of interpolated values obtained by applying locally the selected interpolators.