Spectral edge detection in two dimensions using wavefronts

TL;DR

This paper introduces a multidimensional spectral algorithm to detect wavefronts, capturing both the location and orientation of singularities in functions, improving edge detection in image processing and PDE solutions.

Contribution

It presents a novel wavefront detection method from spectral data that encodes both position and orientation of singularities in multiple dimensions.

Findings

Successfully identifies wavefronts in multidimensional spectral data

Differentiates between singular support and wavefront in higher dimensions

Potential applications in MRI segmentation

Abstract

A recurring task in image processing, approximation theory, and the numerical solution of partial differential equations is to reconstruct a piecewise-smooth real-valued function f(x) in multiple dimensions from its truncated Fourier transform (its truncated spectrum). An essential step is edge detection for which a variety of one-dimensional schemes have been developed over the last few decades. Most higher-dimensional edge detection algorithms consist of applying one-dimensional detectors in each component direction in order to recover the locations in R^N where f(x) is singular (the singular support). In this paper, we present a multidimensional algorithm which identifies the wavefront of a function from spectral data. The wavefront of f(x) is the set of points $(x,k) \in R^N \times (S^{N-1} / \{\pm 1\})$ which encode both the location of the singular points of a function and the orientation of the singularities. (Here $S^{N-1}$ denotes the unit sphere in N dimensions.) More precisely, k is the direction of the normal line to the curve or surface of discontinuity at x. Note that the singular support is simply the projection of the wavefront onto its x-component. In one dimension, the wavefront is a subset of R^1 and coincides with the singular support. In higher dimensions, geometry comes into play and they are distinct. We discuss the advantages of wavefront reconstruction and indicate how it can be used for segmentation in magnetic resonance imaging (MRI).