Reducing the number of vertices in a clustered delaunay mesh

摘要

What is disclosed is a system and method for reducing the number of vertices obtained when building an image gamut from a large number of image pixels. In one embodiment, a color image comprising a plurality of image pixels is received and the color data points measured. An n-dimensional Delaunay triangulation is defined wherein vertices of the triangulation comprise clusters of the measured data points. Cluster analysis is performed on the color data points to determine which existing cluster will receive a new point or if a new cluster should be created. The point clusters are added to the triangulation. For point clusters within the triangulation, a representative point is defined which serves as a normal vertex for the point cluster. The vertex is then added into the triangulation and a gamut generated. The gamut can be used to model a gamut of the color image or a color device.

主权项

1. A method for reducing complexity when constructing a color gamut with a Delaunay mesh, the method comprising:
receiving, into a memory, a plurality of color data points of image pixels of a color image; defining an n-dimensional Delaunay mesh of individual point clusters; performing cluster analysis on said received plurality of color data points to determine which new point clusters need to be added to said mesh and which existing point cluster in said mesh is to receive a color data point; restructuring said n-dimensional Delaunay mesh comprising:
for each point cluster in said Delaunay mesh,
determining a representative point to serve as a normal vertex for said cluster; andadding said normal vertex to said Delaunay mesh; generating a color gamut from said Delaunay mesh; storing said generated color gamut to said memory; and using said generated color gamut to model any of: the gamut of said color image or the gamut of a color output device used to render said color image.