Abstract
The application of multispectral imaging technology in color science is becoming more and more extensive, and how to reconstruct high-precision spectral reflectance information is a key issue currently facing. In order to solve the problem of pixel shift between channels in the band-pass filter spectral imaging system and reduce the complexity of multispectral image acquisition, a spectral reconstruction algorithm based on camera response expansion combined with BP neural network is proposed. Taking spectral root mean square error and CIEDE2000 color difference as evaluation indicators, using SG140 color card as training sample and RC24 color card as test sample, and comparing with several current PCA and BP-PCA-filter methods with better reconstruction effect. The experimental results show that when the number of response expansion items is 22, the average spectral root mean square error and CIEDE2000 color difference are reduced to 0.029 and 1.432, respectively. The spectral reconstruction accuracy of the improved method is obviously better than that of PCA and BP-PCA-filter algorithms.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Farkas, et al.: Microscopic and mesoscopic spectral bio-imgaing. In: Proceedings of the Society of Photo Optical Instrumentation Engineering, vol. 2678, pp. 200–206 (1998)
Maitre, et al.: High Quality Imaging in Museum: From Theory to Practice, pp. 30–39 (1997)
Reng, P.Y., Liao, N.F.: Spectral reflectance reconstruction based on multispectral imaging. Opt. Technol. 31(3), 427–433 (2015)
Li, P.: Research on Micro-vein Imaging Technology Based on Multispectral Imaging. Guangdong University of Technology (2015).
Hill, B.: The History of Multispectral Imaging at Aachen University of Technology (2002)
Wang, X., Thomas, J.-B., Hardeberg, J.Y.: A study on the impact of spectral characteristics of filters on multispectral image acquisition. In: 12th Congress of the International Colour Association (AIC), Newcastle, UK, July (2013)
Shen, H.-L., Mou, T.-S., Xin, J.H.: Colorimetric characterization of scanner by measures of perceptual color error. J. Electron. Imaging 15(4), 41204–41209 (2006)
Li, Y., Shi, B.M.: Research on prediction of coal and gas outburst based on Bayesian regulation BP artificial neural network. In: Working Condition Automation, (2), pp. 1–4 (2009)
Sun, B., Yang, W.P., Hou, D.D., et al.: Design and research of multispectral imaging system based on CMOS. J. Yunnan Normal Univ. (Nat. Sci. Edit.) 28(2), 32–36 (2008)
Liang, J.X., Wan, X.X.: Research on spectral reconstruction of single RGB image of color digital camera. Acta Opt. Sin. 37(09), 370–377 (2017)
Luo, M.R., Cui, G., Rigg, B.: The development of the CIE 2000 colour-difference formula: CIEDE2000. Color Res. Appl. 26(5):340–350 (2001)
Xiao, Y., Ma, X.C., Wang, D.D., Wang, X.H.: Spectral reflectance reconstruction algorithm based on extended three-channel response term 45(06):696–700 (2019)
Acknowledgements
This work is supported by the Yunnan University Undergraduate Education and Teaching Reform Project (JG2018056), Yunnan Normal University Graduate Core Course Construction Project (YH2018-C04), and National University Student Innovation Project (202010681050).
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Wang, D., Feng, J., Zhang, F., Li, X., Zhuo, W., Yu, X. (2021). A Study of Digital Camera Spectral Reconstruction Based on BP Neural Networks and Polynomial Expansions. In: Jain, L.C., Kountchev, R., Shi, J. (eds) 3D Imaging Technologies—Multi-dimensional Signal Processing and Deep Learning. Smart Innovation, Systems and Technologies, vol 234. Springer, Singapore. https://doi.org/10.1007/978-981-16-3391-1_1
Download citation
DOI: https://doi.org/10.1007/978-981-16-3391-1_1
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-16-3390-4
Online ISBN: 978-981-16-3391-1
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)