Essay on Statistical Shape Modelling Of Gaussian Derivatives Used

1336 Words Mar 10th, 2015 6 Pages
3 STATISTICAL SHAPE MODELLING
This section briefly reviews the ASM segmentation scheme. Several comparable approaches are found in the literature. ASMs have been used for several segmentation tasks in medical images. Shapes and objects have been modeled by landmarks, finite-element methods and Fourier descriptors and by expansion in. While there are differences, the general layout of these schemes is similar in that there are: 1) a shape model that ensures that the segmentation can only produce plausible shapes; 2) a gray-level appearance model that ensures that the segmentation places the object at a location where the image structure around the border or within the object is similar to what is expected from the training images. In previous works filter bank of Gaussian derivatives used.
Subsequently a statistical analysis is performed to learn which descriptors are the most informative at each resolution, and at each landmark.and 3) an algorithm for fitting the model by minimizing some cost function.
3.1 Active Shape Models
An object is described by n points, referred to as landmark points. The landmark points are determined in a set of s training images. From these collections of landmark points, a point distribution model is constructed as follows (x1,y1)…(xn,yn). The landmark points are stacked in shape vectors

(2)

(3)…

Related Documents