Shape Characterization via Boundary Distortion

TL;DR

This paper introduces a new shape descriptor based on boundary distortion that is invariant to common transformations and includes a metric for shape comparison, demonstrated through shape retrieval experiments.

Contribution

The paper presents a novel shape descriptor derived from boundary distortion analysis, invariant to rotation, reflection, translation, and scaling, with an associated continuous metric.

Findings

Shape descriptor is invariant to transformations

Proposed metric is continuous and well-defined

Effective shape retrieval demonstrated on two databases

Abstract

In this paper, we derive new shape descriptors based on a directional characterization. The main idea is to study the behavior of the shape neighborhood under family of transformations. We obtain a description invariant with respect to rotation, reflection, translation and scaling. A well-defined metric is then proposed on the associated feature space. We show the continuity of this metric. Some results on shape retrieval are provided on two databases to show the accuracy of the proposed shape metric.