Active shape model

Active shape models (ASMs) are <a href="/facts/Statistical_model/jKkT5ftm">statistical models</a> of the <a href="/facts/Shape/nww0x85d">shape</a> of objects which iteratively deform to fit to an example of the object in a new image, developed by Tim Cootes and Chris Taylor in 1995. The shapes are constrained by the PDM (<a href="/facts/Point_distribution_model/kJUMBtDj">point distribution model</a>) <a href="/facts/Statistical_Shape_Model/KBSm8e7u">Statistical Shape Model</a> to vary only in ways seen in a training set of labelled examples. 
The shape of an object is represented by a set of points (controlled by the shape model). The ASM algorithm aims to match the model to a new image.

The ASM works by alternating the following steps:

<ul><li>Generate a suggested shape by looking in the image around each point for a better position for the point. This is commonly done using what is called a "profile model", which looks for strong edges or uses the <a href="/facts/Mahalanobis_distance/UneEsDng">Mahalanobis distance</a> to match a model template for the point.</li>
<li>Conform the suggested shape to the point distribution model, commonly called a "shape model" in this context. The figure to the right shows an example.</li></ul>
The technique has been widely used to analyse images of faces, mechanical assemblies and medical images (in 2D and 3D).
It is closely related to the <a href="/facts/Active_appearance_model/BDzYLnqB">active appearance model</a>. It is also known as a "Smart Snakes" method, since it is an analog to an <a href="/facts/Active_contour_model/FKekSXgm">active contour model</a> which would respect explicit shape constraints.

Active shape model open-in-new

Active shape model