Toward a computational theory of shape: An overview
Generate an AI Snapshot to get a quick, structured summary of this paper.
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
A theory is proposed that unifies the competing elements of shape—parts and protrusions—and a framework for computing them reliably is developed, which has application in areas ranging from robotics to the psychology and physiology of form.
Abstract
Although the shape of objects is a key to their recognition, viable theories for describing shape have been elusive. We propose a theory that unifies the competing elements of shape—parts and protrusions—and we develop a framework for computing them reliably. The framework emerges from introducing conservation laws to computational vision, and has application in areas ranging from robotics to the psychology and physiology of form.
