match_template_dilate#

pyxem.utils.diffraction.match_template_dilate(image, template, template_dilation=2, mode='constant', constant_values=0)[source]#

Matches a template with an image using a window normalized cross-correlation.

This performs very well for image with different background intensities. This is a slower version of the skimage skimage.feature.match_template() but performs better for images with circular variations in background intensity, specifically accounting for an amorphous halo around the diffraction pattern.

Parameters:

  • image (np.array) – Image to be matched

  • template (np.array) – Template to preform the normalized cross-correlation with

  • template_dilation (int) – The number of pixels to dilate the template by for the windowed cross-correlation

  • mode (str) – Padding mode for the image. Options are ‘constant’, ‘edge’, ‘wrap’, ‘reflect’

  • constant_values (int) – Value to pad the image with if mode is ‘constant’

Returns:

response – The windowed cross-correlation of the image and template

Return type:

np.array