Quick start

Quick start#

This is a quick start example of how to use the SDeconv library. This section supposes you to know the principles of deconvolution. If it is not the case, please refer to the References.

Input images#

Input images are 2D or 3D gray scaled images. 2D images are represented as torch tensors with the following columns ordering [Y, X] and 3D images are represented with torch tensors with [Z, Y, X] columns ordering

Sample images can be loaded using the data module:

from sdeconv import data

image = data.celegans()

Deconvolution using the API#

Bellow is an example how to write a deconvolution script with the API. In this example, we run the Wiener deconvolution algorithm:

from sdeconv import data
from sdeconv.api import SDeconvAPI
import matplotlib.pyplot as plt

# Instantiate the API
api = SDeconvAPI()

# Load image
image = data.celegans()

# Generate a PSF
psf = api.generate_psf('SPSFGaussian', sigma=[1.5, 1.5], shape=[13, 13])

# Deconvolution with API
image_decon = api.deconvolve(image, "SWiener", plane_by_plane=False, psf=psf, beta=0.005, pad=13)

# Plot the results
plt.figure()
plt.subplot(131)
plt.title('Original')
plt.imshow(image.detach().cpu().numpy(), cmap='gray')
plt.axis('off')

plt.subplot(132)
plt.title('PSF')
plt.imshow(psf.detach().cpu().numpy(), cmap='gray')
plt.axis('off')

plt.subplot(133)
plt.title('Wiener deconvolution')
plt.imshow(image_decon.detach().cpu().numpy(), cmap='gray')
plt.axis('off')

plt.show()

The advantage of using the API is that it implements several strategies to deconvolve a 3D or 3D+t image either plane by plane then frame by frame, or frame by frame in 3D.

Deconvolution using the library classes#

When we need only one method, the easiest way may be to call direclty the class that implements the deconvolution algorithm:

import matplotlib.pyplot as plt
from sdeconv.data import celegans
from sdeconv.psfs import SPSFGaussian
from sdeconv.deconv import SWiener

# Load a 2D sample
image = celegans()

# Generate a 2D PSF
psf_generator = SPSFGaussian((1.5, 1.5), (13, 13))
psf = psf_generator()

# Apply Wiener filter
wiener = SWiener(psf, beta=0.005, pad=13)
out_image = wiener(image)

# Display results
plt.figure()
plt.title('PSF')
plt.imshow(psf.detach().numpy(), cmap='gray')

plt.figure()
plt.title('C. elegans original')
plt.imshow(image.detach().numpy(), cmap='gray')

plt.figure()
plt.title('C. elegans Wiener')
plt.imshow(out_image.detach().numpy(), cmap='gray')

plt.show()

Please refer to Modules for more details on the interfaces and the list of available PSFs and deconvolution methods.

Contents