Hostname: page-component-cd9895bd7-q99xh Total loading time: 0 Render date: 2024-12-23T17:34:26.174Z Has data issue: false hasContentIssue false
  • English
  • Français

A geometrical derivation of the shape density

Published online by Cambridge University Press:  01 July 2016

Colin R. Goodall  and
Kanti V. Mardia
Colin R. Goodall*
Affiliation:
Princeton University
Kanti V. Mardia*
Affiliation:
Leeds University
*
∗Postal address: Department of Civil Engineering and Operations Research, School of Engineering and Applied Science, Princeton University, Princeton, NJ 08544, USA.
∗∗Postal address: Department of Statistics, School of Mathematics, University of Leeds, Leeds LS2 9JT, UK.
Get access Rights & Permissions [Opens in a new window]

Abstract

The density for the shapes of random configurations of N independent Gaussian-distributed landmarks in the plane with unequal means was first derived by Mardia and Dryden (1989a). Kendall (1984), (1989) describes a hierarchy of spaces for landmarks, including Euclidean figure space containing the original configuration, preform space (with location removed), preshape space (with location and scale removed), and shape space. We derive the joint density of the landmark points in each of these intermediate spaces, culminating in confirmation of the Mardia–Dryden result in shape space. This three-step derivation is an appealing alternative to the single-step original derivation, and also provides strong geometrical motivation and insight into Kendall's hierarchy. Preform space and preshape space are respectively Euclidean space with dimension 2(N–1) and the sphere in that space, and thus the first two steps are reasonably familiar. The third step, from preshape space to shape space, is more interesting. The quotient by the rotation group partitions the preshape sphere into equivalence classes of preshapes with the same shape. We introduce a canonical system of preshape coordinates that include 2(N–2) polar coordinates for shape and one coordinate for rotation. Integration over the rotation coordinate gives the Mardia–Dryden result. However, the usual geometrical intuition fails because the set of preshapes keeping the rotation coordinate (however chosen) fixed is not an integrable manifold. We characterize the geometry of the quotient operation through the relationships between distances in preshape space and distances among the corresponding shapes.

Keywords

SIZE-AND-SHAPEPRESHAPELANDMARKSOFFSET NORMALREPEATED NORMAL INTEGRALRIEMANNIAN SUBMERSION
Type
Research Article
Information
Advances in Applied Probability , Volume 23 , Issue 3 , September 1991 , pp. 496 - 514
Copyright
Copyright © Applied Probability Trust 1991 

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

Abramowitz, M. and Stegun, I. A. (1972) Handbook of Mathematical Functions. Applied Mathematics Series 55, National Bureau of Standards, United States. Dover, New York.Google Scholar
Bookstein, F. L. (1984) A statistical method for biological shape comparisons. J. Theoret. Biol. 107, 475–520.Google Scholar
Bookstein, F. L. (1986) Size and shape spaces for landmark data in two dimensions (with discussion). Statist. Sci. 1, 181–242.Google Scholar
Crampin, M. and Pirani, F. A. E. (1986) Applicable Differential Geometry. Cambridge University Press, London Math. Soc. Lecture Note Series No. 59.Google Scholar
Dryden, I. L. and Mardia, K. V. (1991) General shape distributions in a plane. Adv. Appl. Prob. 23, 259–276.Google Scholar
Goodall, C. R. (1983) The Statistical Analysis of Growth in Two Dimensions. Doctoral dissertation, Department of Statistics, Harvard University.Google Scholar
Goodall, C. R. (1984) The growth of a two-dimensional figure: strain crosses and confidence regions. American Statistical Association: 1984 Proceedings of the Statistical Computing Section , 165–169.Google Scholar
Goodall, C. R. (1990) Regression models for shape and form. Manuscript.Google Scholar
Goodall, C. R. (1991) Procrustes methods in the statistical analysis of shape (with discussion). J. R. Statist. Soc. 53, 285–339.Google Scholar
Goodall, C. R. and Bose, A. (1987). Procrustes techniques for the analysis of shape change. Proc. Symp. Interface between Computer Science and Statistics , Philadelphia, 86–92.Google Scholar
Goodall, C. R. and Green, P. B. (1986) Quantitative analysis of surface growth. Botanical Gazette 147, 1–15.Google Scholar
Goodall, C. R. and Mardia, K. V. (1990a) The noncentral Bartlett decomposition, and shape densities. J. Multivariate Anal. To appear.Google Scholar
Goodall, C. R. and Mardia, K. V. (1990b) Multivariate aspects of shape theory. Manuscript.Google Scholar
Kendall, D. G. (1984) Shape manifolds, Procrustean metrics, and complex projective spaces. Bull. London Math. Soc. 16, 81–121.Google Scholar
Kendall, D. G. (1989) A survey of the statistical theory of shape. With discussion. Statist. Sci. 4, 87–120.Google Scholar
Kendall, D. G. and Le, H. L. (1990) Riemannian structure of Euclidean shape spaces. Ann. Statist. To appear.Google Scholar
Mardia, K. V. (1972) Statistics of Directional Data. Academic Press, London.Google Scholar
Mardia, K. V. (1989a) Some contributions to shape analysis. Comment on D. G. Kendall's ‘A survey of the statistical theory of shape’. Statist. Sci. 4, 108–111.CrossRefGoogle Scholar
Mardia, K. V. (1989b) Shape analysis of triangles through directional techniques J. R. Statist. Soc. 51, 449–458.Google Scholar
Mardia, K. V. and Dryden, I. L. (1989a) Shape distributions for landmark data. Adv. Appl. Prob. 21, 742–755.Google Scholar
Mardia, K. V. and Dryden, I. L. (1989b) The statistical analysis of shape data. Biometrika 76, 271–282.Google Scholar
Mardia, K. V., Kent, J. T. and Bibby, J. M. (1979) Multivariate Analysis. Academic Press, London.Google Scholar
Small, C. G. (1988) Techniques of shape analysis on sets of points. Internat. Statist. Rev. 56, 243–257.Google Scholar