Exact Point Location and Region-Filling with Quasi-Linear Complexity

In this work we present an extended version of a previous short publication in which we had improved our region filling technique. This technique is correct at locating points and at filling the interior of any connected digital picture. Given any such black- and-white picture, our method will work only with integers to find its exact set of interior points. It then stores these points in a locating matrix, and displays them in yellow colour for visualisation. Our technique is fully programmed in GNU Octave with only 193 lines of source code, which one can apply in computer graphics or image processing straightforwardly.

Autorentext

Valério Ramos Batista is a full professor in computational geometry at the Federal University of ABC, St André, Brazil. Francisco de Assis Zampirolli is a professor at the Federal University of ABC and a researcher in computer vision. Henrique Martins Botelho works in Cryptography, Blockchain and Web3.