Fingerprints, lemniscates and quadratic differentials

TL;DR

This paper explores the mathematical theory behind shape recognition using fingerprints of Jordan curves, linking quadratic differentials and functional equations to extend previous results in shape analysis.

Contribution

It develops a unified approach connecting fingerprints of Jordan curves with quadratic differentials and functional equations, generalizing prior specific cases.

Findings

Established a relation between fingerprints and quadratic differentials.

Unified previous results as special cases of a more general theory.

Extended shape recognition methods using complex analysis techniques.

Abstract

We discuss some aspects of the theory of recognition of two-dimensional shapes by means of fingerprints of Jordan curves. An interesting approach to problems on shape recognition suggested by P.~Ebenfelt, D.~Khavinson, and H.~Shapiro and extended further by M.~Younsi reveals the fact that the fingerprints of polynomial lemniscates and, more generally, fingerprints of rational lemniscates can be obtained as solutions to certain functional equations involving Blaschke products. Our main goal here is to develop an approach which relates fingerprints of Jordan curves composed of arcs of trajectories and orthogonal trajectories of certain quadratic differentials with solutions of functional equations involving pullbacks of these quadratic differentials under appropriate Riemann mapping functions. In particular, we show that the previous results of P.~Ebenfelt, D.~Khavinson, and H.~Shapiro and the recent results of M.~Younsi follow from our more general theorems as special cases.