enzpy

Implementation of the extended Nijboer-Zernike (ENZ) theory for Python.

This toolbox can be used to compute the point-spread function (PSF) using the scalar ENZ theory, see [ENZ], [J2002], [B2008], and [H2010]. It also contains code to fit the phase and the generalised pupil function using real- and complex-valued Zernike polynomials, see [A2015].

Main Features

real- and complex-valued Zernike polynomials
complex point-spread function computation
multi-threaded computation of the V_n^m terms (Eq.(2.48) in [B2008])
routines to fit and evaluate the phase and the generalised pupil function
load/save functions for each object
numerous examples & documentation
ENZPL algorithm example for phase retrieval

Requirements

Python 3
pthreads (tested on Linux and Mac OS X Yosemite)
Numpy
h5py
[optional] matplotlib
[optional] PyQt5
[optional] CVXOPT

Installation

Linux

Make sure you have installed the packages in Requirements.

$ git clone https://github.com/jacopoantonello/enzpy.git
$ cd enzpy
$ sudo python setup.py install

Mac OS X

The easiest way to use this toolbox is to install Anaconda for Python 3, which includes all the necessary packages in Requirements, except for PyQt5 and CVXOPT. Once you have installed Anaconda, create an environment:

$ conda create -n py3 python=3 anaconda
$ source activate py3

and install `enzpy`:

$ git clone https://github.com/jacopoantonello/enzpy.git
$ cd enzpy
$ python setup.py install

PyQt5 is necessary to run the examples with a graphical interface: alpha_abs.py and alpha_abs_qt.py.

Examples

After installing enzpy, you can run the examples located in examples/ (some screenshots are here):

through_focus_intensity.py is taken from [ENZ], and computes the intensity as a function of the radial coordinate and the defocus parameter.
psf_plot.py plots a diffraction-limited PSF at different defocus planes.
phase_plot.py plots the first 10 real-valued Zernike polynomials.
fit_phase.py estimates a vector of real-valued Zernike coefficients from a phase grid by taking inner products numerically.
fit_gpf estimates a vector of real-valued Zernike coefficients from a phase grid by taking inner products numerically. The coefficients can be used to approximate the generalised pupil function.
beta_abs.py and beta_abs_qt.py plot the point-spread function that corresponds to a given complex-valued Zernike analysis of the generalised pupil function. The coefficients can be adjusted using the command line (beta_abs.py) or a Qt widget (beta_abs_qt.py).
alpha_abs.py and alpha_abs_qt.py plot the point-spread function that corresponds to a given real-valued Zernike analysis of the phase aberration function. The coefficients can be adjusted using the command line (alpha_abs.py) or a Qt widget (alpha_abs_qt.py).
enzpl/run contains an example of the ENZPL algorithm, which uses PhaseLift (see [C2013]) and the ENZ theory to correct a random aberration.

Alternatively, you can execute the consistency tests:

$ cd tests
$ nosetests -v -x --pdb *.py

vascotenner/enzpy