grains.segmentation.Segmentation¶

class grains.segmentation.Segmentation(image_location, save_location=None, interactive_mode=True)[source]¶

Segmentation of grain-based microstructures

original_image¶

Matrix representing the initial, unprocessed image.

save_location¶

Directory where the processed images are saved

__init__(image_location, save_location=None, interactive_mode=True)[source]¶

Initialize the class with file paths and with some options

Parameters

image_location (str) – Path to the image to be segmented, file extension included.
save_location (str, optional) – Path to directory where images will be outputted. If not given, the same directory is used where the input image is loaded from.
interactive_mode (bool, optional) – When True, images of each image manipulation step are plotted and details are shown in the console. Default is False.

Returns

None.

Methods

`__init__`(image_location[, save_location, …])	Initialize the class with file paths and with some options
`create_skeleton`(boundary_image)	Use thinning on the grain boundary image to obtain a single-pixel wide skeleton.
`filter_image`(window_size[, image_matrix])	Median filtering on an image.
`find_grain_boundaries`(segmented_image)	Find the grain boundaries.
`initial_segmentation`(*args)	Perform the quick shift superpixel segmentation on an image.
`merge_clusters`(segmented_image[, threshold])	Merge tiny superpixel clusters.
`save_array`(filename, array)	Save an image as a numpy array.
`save_image`(filename, array[, is_label_image])	Save an image as a numpy array.
`watershed_segmentation`(skeleton)	Watershed segmentation of a granular microstructure.