lentil.jitter#

lentil.jitter(img, scale, pixelscale=1, oversample=1)[source]#

Apply image jitter via convolution.

Parameters:

  • img (array_like) – Input image

  • scale (float) –

    1-sigma jitter motion

    If pixelscale = 1, scale should give jitter in terms of fractional pixels. If pixelscale is the physical dimension of one pixel, scale should give jitter in terms of the same units.

  • pixelscale (float, optional) – Pixel size. If pixelscale = 1 (default), the scale parameter provided when Jitter is called should give jitter in terms of fractional pixels. If pixelscale is the physical dimension of one pixel, the ``scale parameter provided when Jitter is called should give jitter in terms of the same units.

  • oversample (int, optional) – Oversampling factor of img. Default is 1.

Returns:

out – Image with jitter applied.

Return type:

ndarray

Examples

Apply 2 pixels jitter to a 5x oversmapled PSF:

>>> import lentil
>>> import matplotlib.pyplot as plt
>>> psf = ...  # PSF calculation details omitted
>>> psf_jitter = lentil.jitter(psf, scale=2, oversample=5)
>>> plt.subplot(121), plt.imshow(psf)
>>> plt.subplot(122), plt.imshow(psf_jitter)
..plot:: _img/python/jitter.py
scale:

50

Apply 10 um jitter to a 5x oversampled PSF. Note that because we are specifying jitter in terms of linear distance on the focal plane, we must provide the detector pixelscale when applying the Jitter. We also provide the oversampling factor, ensuring the convolution kernel is appropriately sized:

>>> import lentil
>>> import matplotlib.pyplot as plt
>>> psf = ...  # PSF calculation details omitted
>>> psf_jitter = lentil.jitter(psf,
...                            scale=10e-6,
...                            pixelscale=5e-6,
...                            oversample=5)
>>> plt.subplot(121), plt.imshow(psf)
>>> plt.subplot(122), plt.imshow(psf_jitter)
../../_images/jitter.png

References

[1] https://en.wikipedia.org/wiki/Convolution_theorem

On this page