13 May 2017 Improving reflectance reconstruction from tristimulus values by adaptively combining colorimetric and reflectance similarities
Author Affiliations +
Abstract
The use of spectral reflectance as fundamental color information finds application in diverse fields related to imaging. Many approaches use training sets to train the algorithm used for color classification. In this context, we note that the modification of training sets obviously impacts the accuracy of reflectance reconstruction based on classical reflectance reconstruction methods. Different modifying criteria are not always consistent with each other, since they have different emphases; spectral reflectance similarity focuses on the deviation of reconstructed reflectance, whereas colorimetric similarity emphasizes human perception. We present a method to improve the accuracy of the reconstructed spectral reflectance by adaptively combining colorimetric and spectral reflectance similarities. The different exponential factors of the weighting coefficients were investigated. The spectral reflectance reconstructed by the proposed method exhibits considerable improvements in terms of the root-mean-square error and goodness-of-fit coefficient of the spectral reflectance errors as well as color differences under different illuminants. Our method is applicable to diverse areas such as textiles, printing, art, and other industries.
© 2017 Society of Photo-Optical Instrumentation Engineers (SPIE)
Bin Cao, Bin Cao, Ningfang Liao, Ningfang Liao, Yasheng Li, Yasheng Li, Haobo Cheng, Haobo Cheng, } "Improving reflectance reconstruction from tristimulus values by adaptively combining colorimetric and reflectance similarities," Optical Engineering 56(5), 053104 (13 May 2017). https://doi.org/10.1117/1.OE.56.5.053104 . Submission: Received: 19 October 2016; Accepted: 19 April 2017
Received: 19 October 2016; Accepted: 19 April 2017; Published: 13 May 2017
JOURNAL ARTICLE
6 PAGES


SHARE
Back to Top