Image interpolation method with decision mixing

摘要

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.

主权项

1. An image interpolation method, comprising:
receiving an input signal having pixel values on an input grid; determining a respective loss value for each of a plurality of candidate interpolators, for a pixel of an output grid which is not located on the input grid, by analyzing the pixel values of the input grid along a direction associated with the candidate interpolator; selecting a fallback interpolator and a plurality of non-fallback interpolators from the plurality of candidate interpolators for said pixel of the output grid based on a comparison of the determined respective loss values for the plurality of candidate interpolators, wherein the fallback interpolator is the interpolator with a lowest loss value among loss values of all of the plurality of candidate interpolators with different directions than directions associated with the selected plurality of non-fallback interpolators; and determining an output pixel value for said pixel of the output grid as a weighted sum of a plurality of interpolated values respectively obtained by applying locally the selected fallback interpolator and a plurality of non-fallback interpolators to pixel values of the input grid, wherein a sum of the plurality of interpolated values is weighted by non-zero coefficients.