光譜晶片完成時，將訊息從像素空間轉換至光譜空間的轉換稱為波長校正，波
長校正需要多組光譜峰值之像素位置與其對應波長。在人為校正的情況下，經常僅
能找到粗略的波峰位置，且計算時耗時許久，因此有不精確、效率低、擬合錯誤這
三個致命的缺點。

在單一筆已知數據情況下要找到光譜中精確的每一個波峰並不困難，本研究提
出一個自動化通用型波長校正方法，對於未知效能之光譜晶片校正仍有良好的效果，
並對於所有校正流程皆提出對應方式與完整的銜接。輸入影像採用橫向掃描式找到
準確ROI，接著Auto Scaling 將光譜最大強度調整至指定強度，解決光譜過曝與過暗
問題。在光譜輸入環境調整完成後，將光譜分區使得每一區域中僅有一波峰，接著
使用高斯擬合或勞倫茲擬合找出精準波峰位置。使用汞氬燈與雷射雙光源校正，結
合雙光源的特徵峰像素位置與絕對波長進行多項式擬合得出空間轉換方程式。本研
究實際校正後之晶片進行空間轉換後波長與標準波長平均誤差低至3nm 內。

When the spectral chip is completed, the conversion of the information from the pixel space to the spectral space is called wavelength correction. Wavelength correction requires multiple pixel position and wavelength sets of peak. In the case of artificial correction, only a rough peak position can be found, and the calculation takes a long time, so there are three fatal shortcomings: inaccuracy, low efficiency, and fitting errors.

It is not difficult to find the precise peaks in the spectrum when the data is unchanged. This research proposes an automated general-purpose wavelength calibration method, which still has a good effect on the calibration of spectrum chips with unknown performance. And for all calibration procedures, corresponding methods and complete connections are proposed.The input image uses horizontal scanning to find the accurate ROI, and then Auto Scaling adjusts the maximum intensity of the spectrum to the specified intensity to solve the problem of over-exposure and very-low-intensity of the spectrum. After the adjustment is completed, the spectrum is divided so that there is only one peak in each region, and then Gaussian fitting or Lorenz fitting is used to find the precise peak position. Use the Hg-Ar lamp and laser multi-light source calibration, combine the characteristic peak pixel positions of the multi-light sources and their absolute wavelengths to perform polynomial fitting to obtain the spatial conversion equation. The RMSE between the converted wavelength of the
corrected chip and the standard wavelength is as low as 3nm.

[1] R. Mabrouki, B. Khaddoumi, and M. Sayadi, “R peak detection in electrocardiogram
signal based on a combination between empirical mode decomposition and hilbert transform,”
in 2014 1st International Conference on Advanced Technologies for Signal and
Image Processing (ATSIP), 2014, pp. 183–187.
[2] S. K. Mukhopadhyay, M. Mitra, and S. Mitra, “ECG signal processing: Lossless compression,
transmission via GSM network and feature extraction using Hilbert transform,”
in 2013 IEEE Point-of-Care Healthcare Technologies (PHT), 2013, pp. 85–88.
[3] C. Bi, L. Li, D. Wang, Z. Wang, and W. Cao, “Research on curve fitting of transmission t
of 2d photonic crystal microcavity,” in 2008 International Workshop on Metamaterials,
2008, pp. 166–168.
[4] C. Li and Y. Li, “Fitting of brillouin spectrum based on labview,” in 2009 5th International
Conference on Wireless Communications, Networking and Mobile Computing,
2009, pp. 1–4.
[5] W. Shan, P. Liu, L. Mu, C. Cao, and G. He, “Hyperspectral image denoising with dual
deep cnn,” IEEE Access, vol. 0, pp. 171 297–171 312, 2019.
[6] L. P. Chang, “Development and application of an algorithm for automated peak detection
and baseline correction for line spectrum.” M.S. thesis, Dept. Automation and
Control., Taiwan Tech Univ., Taipei, Taiwan, 2018, pp. 20–24.
[7] J. C. C. Lo, S. W. R. Lee, X. Guo, and H. Zhao, “Numerical prediction and experimental
validation of multiple phosphor white led spectrum,” in 2016 International Conference
on Electronics Packaging (ICEP), 2016, pp. 57–61.
[8] R. Ferrier, F. Martin, T. Arnoldussen, L. Nunnelley, and M. Burleson, “The determination of transition noise by lorentz electron microscopy,” IEEE Transactions on Magnetics,
vol. 25, no. 5, pp. 3387–3389, 1989.
[9] A. Hepp and C. Baerlocher, “Learned peak shape functions for powder diffraction data,”
Australian Journal of Physics, vol. 41, no. 2, pp. 229–236, 1988.
[10] Y. Fu, Y. Zheng, L. Zhang, and H. Huang, “Spectral reflectance recovery from a single
rgb image,” IEEE Transactions on Computational Imaging, vol. 4, no. 3, pp. 382–394,
2018.
[11] F. Yan, M. Hamit, A. Kutluk, C. Yan, L. Li, W. Yuan, and D. Kong, “Feature extraction
and analysis on x-ray image of xinjiang kazak esophageal cancer by using gray-level
histograms,” in 2013 IEEE International Conference on Medical Imaging Physics and
Engineering. 2nd ed., London, UK: World Scientific Publishing Company, 2013, pp.
61–65.
[12] Y. Chen, Y. Xu, A. J. Mierop, and A. J. P. Theuwissen, “Column-parallel digital correlated
multiple sampling for low-noise cmos image sensors,” IEEE Sensors Journal,
vol. 12, no. 4, pp. 793–799, 2012.
[13] M. Feldman, “Analytic signal representation,” in Hilbert Transform Applications In
Mechanical Vibration. London, UK: John Wiley & Sons, 2011, pp. 9–10.
[14] A. Brown, “Spectral curve fitting for automatic hyperspectral data analysis,” IEEE
Transactions on Geoscience and Remote Sensing, vol. 44, no. 6, pp. 1601–1608, 2006.
[15] V. Jain, M. C. Biesinger, and M. R. Linford, “The gaussian-lorentzian sum, product, and
convolution (voigt) functions in the context of peak fitting x-ray photoelectron spectroscopy
(xps) narrow scans,” Applied Surface Science, vol. 447, pp. 548–553, 2018.
[16] S. Boyd and L. Vandenberghe, “Nonlinear least squares,” in Introduction to Applied Linear Algebra: Vectors, Matrices, and Least Squares. New York, NY, USA: Cambridge
university press, 2018, pp. 381–412.
[17] Z. An and H. Liu, “Achievement of dynamic link library insertion under .net,” in 2010
International Symposium on Intelligence Information Processing and Trusted Computing,
2010, pp. 403–405.
[18] D. Guo, Y. Cheng, S. Zhuo, and T. Sim, “Correcting over-exposure in photographs,” in
2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition,
2010, pp. 515–521.
[19] H. Li, “A novel dual-slope based electrocardiogram peak detection method for wearable
devices,” in 2020 IEEE 3rd International Conference on Information Systems and
Computer Aided Education (ICISCAE), 2020, pp. 327–331.
[20] T. Sasaki, T. Sakamoto, and M. Otsuka, “Absorption peak shape of the lowest vibrational
mode of β-form d-mannitol observed by high frequency accuracy terahertz spectroscopy,”
in 2020 45th International Conference on Infrared, Millimeter, and Terahertz
Waves (IRMMW-THz), 2020, pp. 1–2.
[21] M. N. Vesperinas, “Elements of the theory of diffraction,” in Scattering And Diffraction
In Physical Optics. 2nd ed., London, UK: World Scientific Publishing Company, 2006,
pp. 185–225.
[22] H. Xiaoru and G. Yudong, “Design and implementation of the universal rs232-gpib
interface,” in 2007 8th International Conference on Electronic Measurement and Instruments,
2007, pp. 3–220–3–223.
75