When talking about triangulation, non-mathematicians generally understand it as a geometric method for measuring distances. Roughly speaking, two known points in space can be used to calculate a third via the angles to it. In one or the other Hollywood flick of the genres war or spy movie you have surely come across this.
However, triangulation also refers to the division of a surface into triangles or, more generally, the description of an object by means of polygons. It is used in topology and land surveying, but also in imaging methods of modeling.
How wonderfully this field of mathematics can be applied to photos is shown by the GitHub user RyanH with his program Triangula written in Go, which first roughly splits a given JPG or PNG image into triangles and then refines it further and further via mutations. Among other things, you can specify how many points you want to start with and how many mutations the program should perform. It is also possible to calculate the new image using hexagons instead of classic triangles.