Premium
Textile colour matching using linear and exponential weighted principal component analysis
Author(s) -
Mohtasham Jalil,
Nateri Ali Shams,
Khalili Hale
Publication year - 2012
Publication title -
coloration technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.297
H-Index - 49
eISSN - 1478-4408
pISSN - 1472-3581
DOI - 10.1111/j.1478-4408.2012.00362.x
Subject(s) - principal component analysis , exponential function , mathematics , eigenvalues and eigenvectors , matching (statistics) , pattern recognition (psychology) , artificial intelligence , computer science , statistics , mathematical analysis , physics , quantum mechanics
In this work, linear and exponential weighted principal component analysis techniques based on spectral similarity were employed for the prediction of dye concentration in coloured fabrics, which had been dyed with three component dye mixtures. The matching strategy was based on the equalisation of the first three principal component coordinates of the weighted reflectance curves of the predicted and target sample in a dynamic 3D eigenvector space. The performance of the proposed algorithm was evaluated by the root mean square differences of the reflectance curves and the relative error of the concentration prediction, as well as the metamerism index. The obtained results indicated that the developed exponential weighted principal component analysis method is more accurate than the spectrophotometric method and the simple principal component analysis matching strategy.