Voronoier: from images to Voronoi diagrams

We describe an interactive application for transforming an image into a Voronoi diagram. We combine a variety of methods for generating a spatial point cloud from an input image. In addition, several methods for pruning the spatial point cloud are introduced. These pruning methods can significantly reduce the computation time needed for the transformation. A Voronoi diagram can be constructed from the pruned spatial point cloud using either a naive approach or a Delaunay triangulation. Moreover, an order-k Voronoi diagram can be constructed using this naive approach. We introduce many configuration parameters and we integrate interactivity in the Voronoi diagram by giving users the ability to manually add and remove centroids. To make the application accessible to everyone, we provide a web-based solution by using the Vue.js framework based on the JavaScript programming language. This solution supports the transformation from an image to a Voronoi diagram in a browser, and hence, has the advantage of not being restricted to a certain kind of environment. We illustrate the usefulness of our application by applying it to several images.