[test] OTF estimation without regularization

[kunzaatko]

Test how estimation by simply dividing the Fourier images of the raw image and the ground truth function as an estimate for the OTF. The suspicion is that where-ever there would be small values at high frequencies as expected with these images, the estimate would be overcome by noise and would be very unstable.
$\tilde{d}(\vec{r}) = h(\vec{r}) * d(\vec{r})$ hence $\tilde{D}(\vec{f}) = H(\vec{f})D(\vec{f})$ and an estimate can be made as $\frac{\tilde{D}(\vec{f})}{D(\vec{f})} = H(\vec{f})$.

• Test estimate on new bead data at different locations
• Test with synthetic data where the OTF is known

[kunzaatko]

• Try on an intentionally wrong ground truth