Premium
Alternative mathematical approaches to shade sorting
Author(s) -
Aspland J. R.,
Balasaygun K. D.,
Jarvis J. P.,
Whitaker T. H.
Publication year - 2000
Publication title -
color research and application
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.393
H-Index - 62
eISSN - 1520-6378
pISSN - 0361-2317
DOI - 10.1002/1520-6378(200010)25:5<368::aid-col8>3.0.co;2-f
Subject(s) - cluster analysis , sorting , artificial intelligence , mathematics , set (abstract data type) , computer science , vertex (graph theory) , pattern recognition (psychology) , linkage (software) , computer vision , algorithm , graph , combinatorics , chemistry , biochemistry , gene , programming language
Shade sorting is the process of assigning samples of the same nominal color into groups having no significant color variation. Use of modern spectrophotometers and color measurement technology make it possible to obtain precise color differences between samples. When these color differences are viewed as distances between points, the shade sorting problem is seen to be equivalent to the clustering problem in the mathematical literature. Several mathematical techniques for clustering—complete linkage clustering, vertex labeling, and set covering—are explained and compared for their efficiency when applied to shade sorting. A particular implementation of complete linkage clustering called Clemson Color Clustering (CCC) is found to perform well as compared to the other reviewed methods. © 2000 John Wiley & Sons, Inc. Col Res Appl, 25, 368–375, 2000