ScaleBilinearOpImage.computeValue and InterpolatorBilinear.computeValue have basically the same signature and the same code (only 9 differences in over 200 lines of code, clear sign of copy/paste/adapt).
This means every change to interpolation must be done in two places
For reference, the warp classes are already calling onto the interpolator class instead.
ScaleBilinearOpImage.computeValue and InterpolatorBilinear.computeValue have basically the same signature and the same code (only 9 differences in over 200 lines of code, clear sign of copy/paste/adapt). This means every change to interpolation must be done in two places
For reference, the warp classes are already calling onto the interpolator class instead.