Quick search

Table Of Contents


New in version 1.9.0.

_images/tesselator-filled.png _images/tesselator-debug.png


This is experimental and subject to change as long as this warning notice is present. Only TYPE_POLYGONS is currently supported.

Tesselator is a library for tesselating polygons, based on libtess2. It renders concave filled polygons by first tesselating them into convex polygons. It also supports holes.


First, you need to create a Tesselator object and add contours. The first one is the external contour of your shape and all of the following ones should be holes:

from kivy.graphics.tesselator import Tesselator

tess = Tesselator()
tess.add_contour([0, 0, 200, 0, 200, 200, 0, 200])
tess.add_contour([50, 50, 150, 50, 150, 150, 50, 150])

Second, call the Tesselator.tesselate() method to compute the points. It is possible that the tesselator won’t work. In that case, it can return False:

if not tess.tesselate():
    print "Tesselator didn't work :("

After the tessellation, you have multiple ways to iterate over the result. The best approach is using Tesselator.meshes to get a format directly usable for a Mesh:

for vertices, indices in tess.meshes:

Or, you can get the “raw” result, with just polygons and x/y coordinates with Tesselator.vertices():

for vertices in tess.vertices:
    print "got polygon", vertices
class kivy.graphics.tesselator.Tesselator

Bases: builtins.object

Tesselator class. See module for more informations about the usage.


Add a contour to the tesselator. It can be:

  • a list of [x, y, x2, y2, ...] coordinates
  • a float array: array(“f”, [x, y, x2, y2, ...])
  • any buffer with floats in it.

Returns the number of convex polygon.


Iterate through the result of the tesselate() to give a result that can be easily pushed into Kivy`s Mesh object.

It’s a list of: [[vertices, indices], [vertices, indices], ...]. The vertices in the format [x, y, u, v, x2, y2, u2, v2].

Careful, u/v coordinates are the same as x/y. You are responsible to change them for texture mapping if you need to.

You can create Mesh objects like that:

tess = Tesselator()
# add contours here
for vertices, indices in self.meshes:

Compute all the contours added with add_contour(), and generate polygons.

  • winding_rule (enum) – The winding rule classifies a region as inside if its winding number belongs to the chosen category. Can be one of WINDING_ODD, WINDING_NONZERO, WINDING_POSITIVE, WINDING_NEGATIVE, WINDING_ABS_GEQ_TWO. Defaults to WINDING_ODD.
  • element_type (enum) – The result type, you can generate the polygons with TYPE_POLYGONS, or the contours with TYPE_BOUNDARY_CONTOURS. Defaults to TYPE_POLYGONS.

1 if the tessellation happened, 0 otherwise.

Return type:



Returns the number of vertex generated.

This is the raw result, however, because the Tesselator format the result for you with meshes or vertices per polygon, you’ll have more vertices in the result


Iterate through the result of the tesselate() in order to give only a list of [x, y, x2, y2, ...] polygons.