Abstract
Abstract Shape is defined to be all the geometrical information that is invariant under a set of registration transformations. For example, the registration transformations might be translation, rotation, and scaling. We discuss Kendall's shape space for two‐dimensional sets of landmarks, which is the complex projective space. Probability distributions for planar shapes are reviewed and some approximate distributional results for higher‐dimensional cases are given. Practical shape inference using Procrustes analysis is discussed, and particular emphasis on multivariate analysis in tangent spaces is given. We discuss principal component shape analysis and hypothesis tests, and provide some examples from biology and medicine. Finally, we discuss some applications of shape analysis in image analysis.
Showing the abstract — retrieve the full paper via the Exa API.