A projection algorithm on measures sets

TL;DR

This paper introduces a projection algorithm for probability measures onto sets of Radon measures, motivated by image rendering applications, with convergence analysis and a numerical method demonstrated through artistic drawing examples.

Contribution

It provides a new variational projection framework for measures, establishes conditions for convergence, and develops a numerical algorithm with practical artistic applications.

Findings

Convergence of projections under specific conditions.

Numerical algorithm effectively solves the projection problem.

Applications demonstrated in artistic image synthesis.

Abstract

We consider the problem of projecting a probability measure $\pi$ on a set $\mathcal{M}\_N$ of Radon measures. The projection is defined as a solution of the following variational problem:\begin{equation*}\inf\_{\mu\in \mathcal{M}\_N} \|h\star (\mu - \pi)\|\_2^2,\end{equation*}where $h\in L^2(\Omega)$ is a kernel, $\Omega\subset \R^d$ and $\star$ denotes the convolution operator.To motivate and illustrate our study, we show that this problem arises naturally in various practical image rendering problems such as stippling (representing an image with $N$ dots) or continuous line drawing (representing an image with a continuous line).We provide a necessary and sufficient condition on the sequence $(\mathcal{M}\_N)\_{N\in \N}$ that ensures weak convergence of the projections $(\mu^*\_N)\_{N\in \N}$ to $\pi$.We then provide a numerical algorithm to solve a discretized version of the problem and show several illustrations related to computer-assisted synthesis of artistic paintings/drawings.