Riemannian Geometric Statistics in Medical Image Analysis

11

On shape analysis of functional data

Ruiyi Zhang; Anuj Srivastava    Florida State University, Tallahassee, FL, United States

Abstract

This chapter studies shape analysis of functional data, specifically real-valued functions on intervals and closed curves in ${\mathbb{R}}^2$. This analysis uses a comprehensive elastic Riemannian framework that integrates solutions to the sub-problems of registration, shape comparison and shape analysis, all under the same formulation. Registration implies a dense matching of points across objects when comparing their shapes. This unified framework starts from a square-root transformation of functions, and removes undesired ...