Adding Curvature to Minimum Description Length Shape Models

  • Thodberg H
  • Olafsdottir H
N/ACitations
Citations of this article
26Readers
Mendeley users who have this article in their library.
Get full text

Abstract

The Minimum Description Length ({MDL}) approach to shape modelling seeksa compact description of a set of shapes in terms of the coordinates of markson the shapes. It has been shown that the mark positions resulting from thisoptimisation to a large extent solve the so-called point correspondence problem:How to select points on shapes defined as curves so that the pointscorrespond across a data set. However, this {MDL} approach does not captureimportant shape characteristics related to the curvature of the curves, and occasionallyit places marks in obvious conflict with the human notion of pointcorrespondence. This paper shows how the {MDL} approach can be fine-tunedby adding a term to the cost function expressing the mismatch of curvaturefeatures across the data set. The method is illustrated on silhouettes of adultheads. The {MDL} method is able to solve the point correspondence problemand a classification of the heads into male and female improves dramaticallywhen using the {MDL-}generated marks.

Cite

CITATION STYLE

APA

Thodberg, H. H., & Olafsdottir, H. (2012). Adding Curvature to Minimum Description Length Shape Models (pp. 26.1-26.10). British Machine Vision Association and Society for Pattern Recognition. https://doi.org/10.5244/c.17.26

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free