J. Li, F. Xue and T. Blu, J. Opt. Soc. Am. A, 34(6), pp. 1029-1034, 2017.
This set of codes is a fast approximation of 3D PSF model in fluorescernce microscopy, based on Gibson-Lanni model. See the page for details.
Fast approximation Current approximation can be 498 times faster than the state-of-the-art tool PSF Generator for size 511x511x255.
Note the PSF Generator
uses Java language with mult-threading, while the proposed uses pure Matlab without any compiled routine.
If the parameters of the microscope are not assigned, default setting will be loaded, see MicroscPSF for details.
params.size = [256 256 128];
tic;
PSF = MicroPSF(params);
toc
The last interpolation step is implementated using mex files, which reduce the time cost 50%.
cd Utilities
mex transformation.cpp
params.size = [256 256 128];
params.fastcom = 1;
tic;
PSF = MicroPSF(params);
toc
Mac | |
---|---|
MATLAB R2016a |
MATLAB R2015b |
- (04 May, 2017) Add mex implementation, the speed is further improved 2 times!
- (28 Dec, 2016) Refined basis functions.
- (09 Oct, 2016) Optimized code, the speed has been improved 10 times faster!
-
Matlab code: https://github.com/hijizhou/MicroscPSF-Matlab.
-
ImageJ plugin: https://github.com/hijizhou/MicroscPSF-ImageJ.
-
Icy plugin: https://github.com/hijizhou/MicroscPSF-Icy.
-
Python code (by Dr. Kyle Douglass): http://kmdouglass.github.io/posts/implementing-a-fast-gibson-lanni-psf-solver-in-python.html
Copyright (c) 2017, Jizhou Li, Feng Xue and Thierry Blu
The source code is released under the MIT license.