Region reconstruction with the sphere-of-influence diagram

We describe a simple method for reconstructing regions in the plane from well-distributed point samples. For that, we introduce the sphere-of-influence diagram, a planar diagram extracted from the Delaunay triangulation and the sphere-of-influence graph. The sphere-of-influence diagram is simple to understand and to implement, and supports an intuitive scaling parameter to handle variations in the distribution of samples. We report experiments reconstructing regions from point clouds of varying shape and density. We compare our results with those obtained by alpha-shapes and by cT-shapes. (C) 2022 Elsevier Ltd. All rights reserved.

Automated summary

We present a straightforward approach to reconstructing plane regions from well-distributed point samples. Our method utilizes the sphere-of-influence diagram, which is derived from the Delaunay triangulation and sphere-of-influence graph. Experimental results demonstrate the effectiveness of our method in reconstructing regions from point clouds with varying shapes and densities, achieving competitive performance compared to other algorithms.