Fig. 4.

Voronoi image formation. (A) An image is overlaid on a surface of primer sites. (B) Molecular markers representing different targets (R, G, and B [red, green, and blue]) contact transferred to the polony surface and each covalently linked to a polony barcode. (C) Monte Carlo sampling to determine whether a marker is associated with a given site and if so which target by taking the probability from the RGB value normalized to 1 at the corresponding position in the image. (D) Tallying of markers and empty sites within a polony/Voronoi cell determines the color and brightness of that “pixel.” A subsequent image (D, Lower) is formed by coloring each cell accordingly. (E) Larger-scale reconstruction from scrambled edge data using the Tutte embedding approach with 30,000 polonies. (F) Close-up of E revealing individual Voronoi pixels. Adapted from Da Vinci, Leonardo. 1503–1506. Mona Lisa. Oil on wood. Paris: Musée du Louvre.