HAPPY.image_processing module

HAPPY.image_processing.nan_gaussian(image, sigma)

Apply a gaussian filter to an array with nans.

Parameters

• image (array) – an array with nans

• sigma (float) – σ is the standard deviation of the Gaussian distribution

Returns

gauss – The Gaussian-filtered input image, with nan entries ignored.

Return type

array, same shape as image

HAPPY.image_processing.minimize_grain_contrast(image, sigma)

Minimise grain contrast or uneven lighting.

This is accomplished by dividing the original image by an image with a gaussian blur applied.

Parameters

• image (array) – Image to minimise grain contrast.

• sigma (float) – Sigma value for gaussian blur.

Returns

removed_grains – Output image.

Return type

array, same shape as image

HAPPY.image_processing.simple_threshold(image, scale_um, crop_threshold, threshold, small_obj=None)

Threshold the image, accounting for crop and small features.

Hydrides are assumed to be dark (value below the threshold) in the input image, but are returned as bright (1.0) features in the output, and vice- -versa for the matrix.

Parameters

• removed_grains (array) – image to threshold.

• crop_threshold (array of bool) – Thresholding is only performed within regions labeled False in this array. Values labeled True will be set to np.nan in the output

• scale_um (int) – Scale bar value in microns

• theshold (float) – threshold level.

• small_obj (int, optional) – size of features to be removed and not thresholded in microns

Returns

thres_disp – The thresholded image, with 1.0 in foreground pixels, 0.0 in background pixels, and np.nan in cropped pixels.

Return type

array of float