A0275
Title: Logistic regression models for elastic shape of curves
Authors: Min Ho Cho - Inha University (Korea, South) [presenting]
Abstract: Shape analysis is widely used in many areas, such as computer vision and medical and biological studies. One challenge in analyzing the shape of an object in an image is its invariant property to shape-preserving transformations. To measure the distance or dissimilarity between two shapes, the square-root velocity function (SRVF) representation and the elastic metric are used. Since shapes are inherently high-dimensional in a nonlinear space, a tangent space is adopted at the mean shape and a few Principal Components (PCs) on the linearized space. Classification methods are proposed based on logistic regression using these PCs and tangent vectors with the elastic net penalty. Its performance compared with other model-based methods for shape classification is assessed on the shape of algae in watersheds, as well as simulated data generated by the mixture of von Mises-Fisher distributions.
