A Composition Analysis Method of Mixed Pigments Based on Spectrum Expression and Independent Component Analysis
WANG Gong-ming1, LIU Zhi-yong2
1. Department of Information Art and Design, Tsinghua University, Beijing 100084, China 2. Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190, China
Abstract：Reflectance spectrometry is a common method in composition analysis of mixed pigments. In this method, similarity is used to determine the type of basic pigments that constitute the mixed pigments. But its result may be inaccurate because it is easily influenced by a variety of basic pigments. In this study, a composition analysis method of mixed pigments based on spectrum expression and independent component analysis is proposed, and the composition of mixed pigments can be calculated accurately. First of all, the spectral information of mixed pigments is obtained with spectrometer, and is expressed as the discrete signal. After that, the spectral information of basic pigments is deduced with independent component analysis. Then, the types of basic pigments are determined by calculating the spectrum similarity between the basic pigments and known pigments. Finally, the ratios of basic pigments are obtained by solving the Kubelka-Munk equation system. In addition, the simulated spectrum data of Munsell color card is used to validate this method. The compositions of mixed pigments from three basic pigments are determined under the circumstance of normality and disturbance. And the compositions of mixture from several pigments within the set of eight basic pigments are deduced successfully. The curves of separated pigment spectrums are very similar to the curves of original pigment spectrums. The average similarity is 97.72%, and the maximum one can reach to 99.95%. The calculated ratios of basic pigments close to the original one. It can be seen that this method is suitable for composition analysis of mixed pigments.
王功明1，刘志勇2 . 基于光谱表示和独立成分分析的混合颜料成分分析方法 [J]. 光谱学与光谱分析, 2015, 35(06): 1682-1689.
WANG Gong-ming1, LIU Zhi-yong2 . A Composition Analysis Method of Mixed Pigments Based on Spectrum Expression and Independent Component Analysis . SPECTROSCOPY AND SPECTRAL ANALYSIS, 2015, 35(06): 1682-1689.
[1] Tiziana C, Annamaria G, Marco N. Procedia Chemistry, 2013, 8(1): 45. [2] ZHANG Sheng-ping(张胜平). Based on the Multispectral Imaging of Research on Method of the Color Rejuvenation of Chinese Paintings(基于多光谱图像的古画颜色修复方法研究)，Master Degree Dissertation (硕士论文). Tianjin: Tianjin University(天津: 天津大学), 2010, 8. [3] Richard E B, Rand C. Dynamic Programming. New Jersey: Princeton University Press，1957. 35. [4] Beyel K, Goldstein J, Ranmakrishnan R. Proceedings of the International Conference on Database Theory, 1999. 217. [5] Hyvrinen A, Oja E. Neural computation, 1997, 9(7): 1483. [6] Berns R S. Billmeyer and Saltzman’s Principles of Color Technology(颜色技术原理). Translated by LI Xiao-mei, MA Ru, CHEN Li-rong(李小梅, 马 如, 陈立荣, 译). Beijing: Chemical Industry Press(北京: 化学工业出版社), 2002. 34. [7] Kubelka P, Munk F. Zeitschrift Fur Technische Physik, 1931, 12(11): 593. [8] ZHANG Xian-da, BAO Zheng(张贤达, 保 铮). Chinese Journal of Electronics(电子学报), 2001, 29(12A): 1766. [9] LIN Qiu-hua(林秋华). Novel Methods ofr Encyrpting Images and Speeches Based on Blind Souree Separation (基于盲源分离的图像与语音加密新方法研究), Doctoral Dissertation(博士论文). Dalian: Dalian University of Technology(大连: 大连理工大学), 2006, 20. [10] Sapatnekar S S. IEEE Journal on Emerging and Selected Topics in Circuits and Systems，2011, 1(1): 5. [11] Gillespie A R, Kahle A B, Walker R E. Remote Sensing of Environment, 1986, 20(3): 209. [12] Alsberg B K, Woodward A M, Kell D B. Chemometrics and Intelligent Laboratory Systems, 1997, 37(2): 215. [13] Rice J R. Mathematics of Computation, 1966, 20(94): 325. [14] Householder A S. Journal of the ACM, 1958, 5(4): 339. [15] WANG Wei-hua(王卫华). Research on Blind Source Separation Algorithms and its Application(盲源分离算法及应用研究), Doctoral Dissertation(博士论文). Harbin: Harbin Engineering University(哈尔滨: 哈尔滨工程大学), 2009, 19.
