# Fringe.Py
[](https://badge.fury.io/gh/farhadnkm%2FFringe.Py)
[](https://badge.fury.io/py/fringe)
[](https://zenodo.org/badge/latestdoi/327240077)
[](https://github.com/farhadnkm/Fringe.Py/blob/master/LICENSE)
A coherent propagation and diffraction simulation tool, all written in Python.
In this package, a set of utilities are provided to simulate coherent signal propagation. It is particularly made for free-space optical propagation, diffraction, and holography. However, the tools are compatible with 1D and 2D data structures and can be potentially used for any sort of spatially-coherent signals.
-----
<p align="center">
<img src="examples/images/exports/propagation_anim.gif" width="500">
<p>
## What's inside?
A set of modules are provided to solve the scalar diffraction problem. The codes are GPU-friendly, compatible with well-known computation libraries (Numpy, TensorFlow) and they support batch processing. Using the TensorFlow backend, computations become autograd-compatible and could be mounted on machine learning models. Angular spectrum algorithm is the primary work horse for field propagation though other custom solvers could be used. Aside the built-in Numpy and TensorFlow backends, any computational means could be employed to process tensor operations.
It also includes:
- a simple yet useful data pipeline to load and standardize data. For now, it only supports images.
- Gerchberg-Saxton multi-distance phase recovery algorithm. It can be easily tweaked to support other variations of signal e.g. by wavelength.
## Installation
To install the package, run:
```
python -m pip install fringe
```
Fringe requires ```numpy```, ```tensorflow 2.x```, and ```scikit_image```.
The example files are not included in the package and should be downloaded separately. Also they require ```matplotlib``` to show plots.
## How to Use
1. Import or create data
For images:
```python
import numpy as np
from fringe.utils.io import import_image
from fringe.utils.modifiers import ImageToArray, Normalize, MakeComplex
```
Images need to be standardized, normalized, and casted to complex data type. *Modifiers* are tools made for this purpose which apply these operations on import.
```python
p1 = ImageToArray(bit_depth=16, channel='gray', crop_window=None, dtype='float32')
p2 = Normalize(background=np.ones((512, 512)))
p3 = MakeComplex(set_as='amplitude', phase=0)
obj = import_image("images/squares.png", preprocessor=[p1, p2, p3])
```
2. Propagate
*Solvers* contain propagation algorithms and can be called by *solver.solve*. In particular, angular Spectrum algorithm convolves the input field with a free-space propagtor function which depends on *wavelength λ* (or *wavenumber k=2π/λ*) and distance *z*.
```python
from numpy import pi, abs, angle
from fringe.solvers.AngularSpectrum import AngularSpectrumSolver as AsSolver
solver = AsSolver(shape=obj.shape, dr=1, is_batched=False, padding="same", pad_fill_value=0, backend="Numpy")
rec = solver.solve(hologram, k=2*pi/500e-3, z=-1000)
amp, phase = abs(rec), angle(rec)
ax = pyplot.sublots(2, 1, 1)
ax[0].imshow(abs(obj))
ax[1].imshow(amp)
pyplot.show()
```
<p align="center">
<img src="examples/images/exports/squares_1.png" width="300"> <img src="examples/images/exports/squares_2.png" width="300">
<p>
Example notebooks provide further details with 1D and 2D diffraction, GPU acceleration, batch processing, and phase recovery.
## License
Fringe is released under the MIT license. See LICENSE for details.
=======
History
=======
0.0.1 (2021-02-19)
------------------
* First release on PyPI.
0.0.2 (2021-02-25)
------------------
* Some bugs and errors fixed
0.0.3 (2021-02-25)
------------------
* Minor bugs fixed
0.0.5 (2023-03-15)
------------------
* AngularSpectrum.py
Propagator issue for small pixels is resolved
Near field propagation is now supported
* io.py
skimage warnings io is fixed.
0.0.6 (2023-03-16)
------------------
* AngularSpectrum.py
Bug fixed
Raw data
{
"_id": null,
"home_page": "https://github.com/farhadnkm/fringe",
"name": "fringe",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.6, <4",
"maintainer_email": "",
"keywords": "diffraction,angular spectrum,holography,holograms,phase recovery,",
"author": "Farhad Niknam",
"author_email": "farhad.niknam.em@gmail.com",
"download_url": "",
"platform": null,
"description": "\n\n# Fringe.Py\n\n[](https://badge.fury.io/gh/farhadnkm%2FFringe.Py)\n[](https://badge.fury.io/py/fringe)\n[](https://zenodo.org/badge/latestdoi/327240077)\n[](https://github.com/farhadnkm/Fringe.Py/blob/master/LICENSE)\n\nA coherent propagation and diffraction simulation tool, all written in Python. \n\nIn this package, a set of utilities are provided to simulate coherent signal propagation. It is particularly made for free-space optical propagation, diffraction, and holography. However, the tools are compatible with 1D and 2D data structures and can be potentially used for any sort of spatially-coherent signals.\n\n-----\n\n<p align=\"center\">\n <img src=\"examples/images/exports/propagation_anim.gif\" width=\"500\">\n<p>\n\n \n## What's inside?\n\nA set of modules are provided to solve the scalar diffraction problem. The codes are GPU-friendly, compatible with well-known computation libraries (Numpy, TensorFlow) and they support batch processing. Using the TensorFlow backend, computations become autograd-compatible and could be mounted on machine learning models. Angular spectrum algorithm is the primary work horse for field propagation though other custom solvers could be used. Aside the built-in Numpy and TensorFlow backends, any computational means could be employed to process tensor operations.\n\nIt also includes:\n- a simple yet useful data pipeline to load and standardize data. For now, it only supports images.\n- Gerchberg-Saxton multi-distance phase recovery algorithm. It can be easily tweaked to support other variations of signal e.g. by wavelength.\n\n## Installation\nTo install the package, run:\n\n```\npython -m pip install fringe\n```\n\nFringe requires ```numpy```, ```tensorflow 2.x```, and ```scikit_image```.\nThe example files are not included in the package and should be downloaded separately. Also they require ```matplotlib``` to show plots.\n\n## How to Use\n1. Import or create data\n\nFor images:\n\n```python\nimport numpy as np\nfrom fringe.utils.io import import_image\nfrom fringe.utils.modifiers import ImageToArray, Normalize, MakeComplex\n```\n\nImages need to be standardized, normalized, and casted to complex data type. *Modifiers* are tools made for this purpose which apply these operations on import.\n\n```python\np1 = ImageToArray(bit_depth=16, channel='gray', crop_window=None, dtype='float32')\np2 = Normalize(background=np.ones((512, 512)))\np3 = MakeComplex(set_as='amplitude', phase=0)\n\nobj = import_image(\"images/squares.png\", preprocessor=[p1, p2, p3])\n```\n\n2. Propagate\n\n*Solvers* contain propagation algorithms and can be called by *solver.solve*. In particular, angular Spectrum algorithm convolves the input field with a free-space propagtor function which depends on *wavelength \u00ce\u00bb* (or *wavenumber k=2\u00cf\u20ac/\u00ce\u00bb*) and distance *z*.\n\n```python\nfrom numpy import pi, abs, angle\nfrom fringe.solvers.AngularSpectrum import AngularSpectrumSolver as AsSolver\n\nsolver = AsSolver(shape=obj.shape, dr=1, is_batched=False, padding=\"same\", pad_fill_value=0, backend=\"Numpy\")\nrec = solver.solve(hologram, k=2*pi/500e-3, z=-1000)\namp, phase = abs(rec), angle(rec)\n\nax = pyplot.sublots(2, 1, 1)\nax[0].imshow(abs(obj))\nax[1].imshow(amp)\npyplot.show()\n```\n\n<p align=\"center\">\n <img src=\"examples/images/exports/squares_1.png\" width=\"300\"> <img src=\"examples/images/exports/squares_2.png\" width=\"300\">\n<p>\n\nExample notebooks provide further details with 1D and 2D diffraction, GPU acceleration, batch processing, and phase recovery.\n\n## License\n\nFringe is released under the MIT license. See LICENSE for details.\n\n\n=======\nHistory\n=======\n\n0.0.1 (2021-02-19)\n------------------\n* First release on PyPI.\n\n0.0.2 (2021-02-25)\n------------------\n* Some bugs and errors fixed\n\n0.0.3 (2021-02-25)\n------------------\n* Minor bugs fixed\n\n0.0.5 (2023-03-15)\n------------------\n* AngularSpectrum.py\n Propagator issue for small pixels is resolved\n Near field propagation is now supported\n* io.py\n skimage warnings io is fixed.\n\n0.0.6 (2023-03-16)\n------------------\n* AngularSpectrum.py\n Bug fixed\n",
"bugtrack_url": null,
"license": "MIT license",
"summary": "Python implementation of computational optical propagation and digital holography.",
"version": "1.0.6",
"split_keywords": [
"diffraction",
"angular spectrum",
"holography",
"holograms",
"phase recovery",
""
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "d22125c4d2d12f3c43ce19e3cdf7f5e99f61521673862217b7d74a7652323d15",
"md5": "c1c2c4c6f8afd3b4b66f8255c8812d0c",
"sha256": "ee9d90d83261716dfce4ecff66b07c4873c0c5c67728d2e943dc22c85e0558da"
},
"downloads": -1,
"filename": "fringe-1.0.6-py2.py3-none-any.whl",
"has_sig": false,
"md5_digest": "c1c2c4c6f8afd3b4b66f8255c8812d0c",
"packagetype": "bdist_wheel",
"python_version": "py2.py3",
"requires_python": ">=3.6, <4",
"size": 18145,
"upload_time": "2023-03-16T13:20:32",
"upload_time_iso_8601": "2023-03-16T13:20:32.673394Z",
"url": "https://files.pythonhosted.org/packages/d2/21/25c4d2d12f3c43ce19e3cdf7f5e99f61521673862217b7d74a7652323d15/fringe-1.0.6-py2.py3-none-any.whl",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-03-16 13:20:32",
"github": true,
"gitlab": false,
"bitbucket": false,
"github_user": "farhadnkm",
"github_project": "fringe",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [],
"tox": true,
"lcname": "fringe"
}
JSON
