Spectral reflectance curves for multispectral imaging, combining different techniques and a neural network
C.A. Osorio–Gómezª, E. Mejía–Ospinob, and J.E. Guerrero–Bermúdezª,*
ª Grupo de Óptica y Tratamiento de Señales. Escuela de Física, Facultad de Ciencias,Universidad Industrial de Santander, A.A. 678, Bucaramanga, Colombia.
b Laboratorio de Espectroscopía Atómica Molecular. Escuela de Química, Facultad de Ciencias, Universidad Industrial de Santander, A.A. 678, Bucaramanga, Colombia.
*Corresponding author: ]]>
Tel./Fax: +57 7 6349069
e–mail:jader@uis.edu.co
Recibido el 5 de enero de 2009
Aceptado el 17 de marzo de 2009
Abstract
In this paper, we present an alternative procedure for the digital reconstruction of spectral reflectance curves of oil painting on canvas using multispectral imaging. The technique is based on a combination of the results obtained by pseudo–inverse, principal component analysis and interpolation; these results are the input to a feed–forward back propagation neural network fittin the values of the curves to a target obtained using a spectrophotometer Shimadzu UV2401. Goodness–of–Fit Coefficien (GFC), absolute mean error (ABE) and spectral Root Mean Squared error (RMS) are the metrics used to evaluate the performance of the procedure proposed.
Keywords: Artificia neural network; spectral reflectance curves; multispectral imaging; curve fitting.
]]> Resumen
Se presenta un procedimiento alternativo para la reconstrucción numérica de curvas de reflectancia espectral de pinturas de óleo sobre lienzo. La técnica se basa en la combinación no lineal de los espectros de reflectancia reconstruidos por procedimientos lineales, tales como: pseudoinversa, análisis en componentes principales e interpolación. Estos espectros constituyen la entrada a una red neuronal artificia entrenada mediante el algoritmo backpropagation. La red neuronal ajusta sus pesos y umbrales de acuerdo a los espectros patrones, obtenidos mediante un espectro fotometro Shimadzu UV2401. Para evaluar el desempeño del procedimiento se utilizaron las métricas: cálculo del error espectral cuadrático medio, el coeficiente de buen ajuste y el error medio absoluto.
Descriptores: Espectros de reflectancia; ajuste de curvas; imágenes multiespectrales; redes neuronales artificiales.
PACS: 42.66.Ne; 42.66.Qg; 07.57.–c; 07.05.Mh; 02.70.Hm.
DESCARGAR ARTÍCULO EN FORMATO PDF
Acknowledgment
The authors would like to thank the VIE–UIS (project 5162), for financin this work.
]]>References
1. A. Ribés, Multispectral analysis and spectral reflectance reconstruction of art paintings (Ph.D. dissertation, Ecole Nationale Superieure des Telecommunications", Paris, 2003). [ Links ]
2. J.Y. Hardeberg, Acquisition and reproduction of colour images colorimetric and multispectral approaches (Ph.D. dissertation, Ecole Nationale Supérieure des Telecommunications", Paris, 1999). [ Links ]
3. H.J. Trussell, E. Saber, and M. Vhrel, IEEE Signal Processing Magazine (2005) 14. [ Links ]
4. J.L. Nieves, J.Hernández, E. Valero, and J. Romero, Appl. Opt. 43 (2004) 1880. [ Links ]
5. A. Ribés and F. Schmitt, Pattern Recognition Letter 24 (2003) 1691. [ Links ]
6. H.L. Shen, J.H. Xin, and S.J. Shao. (2007, April) "Improved reflectance reconstruction for multispectral imaging by combining differents techniques," Optics Express (online), 15(7), pp 5531–5536. Available : http://www.opticsexpress.org [ Links ]
7. G. Sharma, Digital Color Imaging Handbook (CRC PRESS Ed., 844 p, New York, 2003). [ Links ]
8. H. Haneishi et al., Appl. Opt. 39 (2000) 6621. [ Links ]
9. J. Conde, H. Haneishi, M. Yamaguchi, N. Ohyama, and J. Baez. Rev. Mex. Fis. 50 (2004) 484. [ Links ]
10. S. Tominaga, spie's oe magazine, (2003) 24. [ Links ]
11. M. Vilaseca, J. Pujol, and M. Arjona Appl. Opt. 42 (2003) 1789. [ Links ]
12. J.H. Mathews, K.D. Fink. Métodos Numéricos con Matlab 3th ed. (Prentice Hall Ed, Madrid, 2000) p. 304–305. [ Links ]
13. Y.H. Hu, J. Hwang, Handbook of Neural Network Signal Processing (CRC PRESS Ed., Boca Raton, 2002). p. 384 [ Links ]
14. H. Demuth, M. Beale, and M. Hagan, Neural Network Toolbox. For use with Matlab (Version 4, The Mathworks, Natick MA., 2005) p. 808. [ Links ]
15. J.A. Freeman and D.V. Skapura, Neural Networks. Algorithms, Applications and Programming Techniques (Addison–Wesley Ed., Massachusetts, 1991) p. 401. [ Links ]
16. N.A. Arias, Sistema de visión artificial para la estimación del peso del ganado bovino (Tesis de Maestría en Física, Universidad Industrial de Santander. Bucaramanga 2004). [ Links ]
17. P.D. Burns and R.S. Berns, "Quantization in multispectral color image adquisition," in Proc. IS & T /SID Seventh color imaging conference: Color science, systems and applications, IS & T (1999) p. 32. [ Links ]
18. P. Carcagni, et al., Optics and Laser in Engineering 45 (2006) 360. [ Links ]
19. H.G. Volz, Industrial Color Testing: Fundamental and Techniques 2nd (Ed. Wiley–VCH, New York, 2001) p. 243. [ Links ]
20. L. G. Valdivieso, J.E. Guerrero, "Numerical Reconstruction of Spectral Reflectanc Curves of Oil Painting on Canvas," in Proc. XII STSIVA, Barranquilla, 2007. Edición electrónica: UNINORTE y Rama estudiantil IEEE Colombia (Student branch of IEEE Colombia). [ Links ] ]]>