Morphological skeleton

In digital image processing, morphological skeleton is a skeleton (or medial axis) representation of a shape or binary image, computed by means of morphological operators.

Morphological skeletons are of two kinds:

• Those defined by means of morphological openings, from which the original shape can be reconstructed,
• Those computed by means of the hit-or-miss transform, which preserve the shape's topology.

In (Lantuéjoul 1977),^[1] Lantuéjoul derived the following morphological formula for the skeleton of a continuous binary image ${\displaystyle X\subset \mathbb {R} ^{2}}$:

${\displaystyle S(X)=\bigcup _{\rho >0}\bigcap _{\mu >0}\left[(X\ominus \rho B)-(X\ominus \rho B)\circ \mu {\overline {B}}\right]}$,

where ${\displaystyle \ominus }$ and ${\displaystyle \circ }$ are the morphological erosion and opening, respectively, ${\displaystyle \rho B}$ is an open ball of radius ${\displaystyle \rho }$, and ${\displaystyle {\overline {B}}}$ is the closure of ${\displaystyle B}$.

Let ${\displaystyle \{nB\}}$, ${\displaystyle n=0,1,\ldots }$, be a family of shapes, where B is a structuring element,

${\displaystyle nB=\underbrace {B\oplus \cdots \oplus B} _{n{\mbox{ times}}}}$, and

${\displaystyle 0B=\{o\}}$, where o denotes the origin.

The variable n is called the size of the structuring element.

Lantuéjoul's formula has been discretized as follows. For a discrete binary image ${\displaystyle X\subset \mathbb {Z} ^{2}}$, the skeleton S(X) is the union of the skeleton subsets ${\displaystyle \{S_{n}(X)\}}$, ${\displaystyle n=0,1,\ldots ,N}$, where:

${\displaystyle S_{n}(X)=(X\ominus nB)-(X\ominus nB)\circ B}$.

The original shape X can be reconstructed from the set of skeleton subsets ${\displaystyle \{S_{n}(X)\}}$ as follows:

${\displaystyle X=\bigcup _{n}(S_{n}(X)\oplus nB)}$.

Partial reconstructions can also be performed, leading to opened versions of the original shape:

${\displaystyle \bigcup _{n\geq m}(S_{n}(X)\oplus nB)=X\circ mB}$.

Let ${\displaystyle nB_{z}}$ be the translated version of ${\displaystyle nB}$ to the point z, that is, ${\displaystyle nB_{z}=\{x\in E|x-z\in nB\}}$.

A shape ${\displaystyle nB_{z}}$ centered at z is called a maximal disk in a set A when:

• ${\displaystyle nB_{z}\in A}$, and
• if, for some integer m and some point y, ${\displaystyle nB_{z}\subseteq mB_{y}}$, then ${\displaystyle mB_{y}\not \subseteq A}$.

Each skeleton subset ${\displaystyle S_{n}(X)}$ consists of the centers of all maximal disks of size n.

Morphological Skeletonization can be considered as a controlled erosion process. This involves shrinking the image until the area of interest is 1 pixel wide. This can allow quick and accurate image processing on an otherwise large and memory intensive operation. A great example of using skeletonization on an image is processing fingerprints. This can be quickly accomplished using bwmorph; a built-in Matlab function which will implement the Skeletonization Morphology technique to the image.

The image to the right shows the extent of what skeleton morphology can accomplish. Given a partial image, it is possible to extract a much fuller picture. Properly pre-processing the image with a simple Auto Threshold grayscale to binary converter will give the skeletonization function an easier time thinning. The higher contrast ratio will allow the lines to joined in a more accurate manner. Allowing to properly reconstruct the fingerprint.

skelIm = bwmorph(orIm,'skel',Inf); %Function used to generate Skeletonization Images 

• Image Analysis and Mathematical Morphology by Jean Serra, ISBN 0-12-637240-3 (1982)
• Image Analysis and Mathematical Morphology, Volume 2: Theoretical Advances by Jean Serra, ISBN 0-12-637241-1 (1988)
• An Introduction to Morphological Image Processing by Edward R. Dougherty, ISBN 0-8194-0845-X (1992)
• Ch. Lantuéjoul, "Sur le modèle de Johnson-Mehl généralisé", Internal report of the Centre de Morph. Math., Fontainebleau, France, 1977.
• P. Maragos and R. W. Schafer, Morphological Skeleton Representation and Coding of Binary Images, IEEE Trans. on Acoustics, Speech, and Signal Processing, vol.34, no.5, pp.1228-1244, Oct. 1986.
• Scott E. Umbaugh (2018). Digital Image Processing and Analysis, pp 93-96. CRC Press. ISBN 978-1-4987-6602-9