Morphological Operations¶
This algorithm computes morphological operations on images.
Input¶
The algorithm requires a single-channel image or image set as input.
Output¶
If the In place flag is disabled, the result will be added as new image to the data model. Otherwise, the processing will be performed in-place.
Description¶
This algorithm performs a morphological operation (based on the Operation) on the input. The Pixel size parameter defines the size of the structuring element, which can be thought as a filter kernel defining the actual operation. The larger the size of the structuring element, the stronger the is the morphological change induced on the image. In this algorithm we only support spherical structuring elements.
The supported operations are:
Dilation: on binary images, it expands the foreground region by the radius of the structuring element, while on greyvalue images it highlights brighter regions.Erosion: on binary images, it shrinks the foreground region by the radius of the structuring element, while on greyvalue images it highlights darker regions.Morphological opening: on binary images, it ‘opens’ corners into rounder edges and removes non connected parts which are smaller or equal to the structuring element size. On greyvalue images the effect of this operation can be thought as a dampening of brighter regions in favor of darker ones.Morphological closing: on binary images, it ‘closes’ smooth edges into shaper corners and fills holes in connected parts which are smaller or as large as the structuring element. On greyvalue images the effect of this operation can be thought as highlighting brighter regions and smoothing out darker ones.Skeletonization: reduces a shape to a thinner version that is equidistant from its boundaries while preserving its connectivity.
Notes¶
Different morphological operations are not commutative: opening and closing for instance are a dilation and erosion with the same structuring element performed in inversed order.