Color quantization is a process of choosing a set of K representative colors to approximate the N colors of an image, K<<N, such that the resulting K-color image looks as much like the original N-color image as possible. This paper presents a new color quantization algorithm by efficient statistical computations and nearest-neighbor searching. Different from recursively bipartition, this algorithm directly splits the color space into many voxels, and chooses a color palette which permits the most pixels with the colors corresponding to the voxels that they fall. The algorithm has a good effectiveness.

Date Published: 22 March 1996
PDF: 5 pages
Proc. SPIE 2644, Fourth International Conference on Computer-Aided Design and Computer Graphics, (22 March 1996); doi: 10.1117/12.235489

Published in SPIE Proceedings Vol. 2644:
Fourth International Conference on Computer-Aided Design and Computer Graphics
Shuzi Yang; Ji Zhou; Cheng-Gang Li, Editor(s)