Constrained high-order statistical blind deconvolution of spectral data
A constrained high-order statistical algorithm is proposed to blindly deconvolute the measured spectral data and estimate the response function of the instruments simultaneously. In this algorithm, no prior-knowledge is necessary except a proper length of the unit-impulse response. This length can be easily set to be the width of the narrowest spectral line by observing the measured data. The feasibility of this method has been demonstrated experimentally by the measured Raman and absorption spectral data.
Ziqiang Hu：Institute of Science and Technology for Opto-Electron Information, Yantai University, Yantai 264005
Guiying Wang：Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, Shanghai 201800
Zhizhan Xu：Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, Shanghai 201800
【1】Y. Senga, K. Minami, S. Kawata, and S. Minami, Appl. Opt. 23, 1601 (1984).
【2】R. D. Davies and P. A. Jansson, "Applications to electron spectroscopy for chemical analysis", in Deconvolution of Images and Spectra (2nd edn.) P. A. Jansson (ed.) (Academic Press, San Diego, 1997).
【3】O. Shalvi and E. Weinstein, "Universal Methods for Blind Deconvolution", in Blind Deconvolution S. Haykin (ed.) (Prentice-Hall, Englewood Cliffs, 1994) pp.122---168.
【4】S. B. Engelson, "Raman Spectral of (D+)-Glucopyranose" http://www.models.kvl.dk/users/ ergelsen/specarb/glcb.html.
【5】NASA, "Absorption Spectral Data of Cr:LiSAF Crystal" http://aesd.larc.nasa.gov/gl/laser/spectra/spectra.htm.
Jinghe Yuan, Ziqiang Hu, Guiying Wang, Zhizhan Xu, "Constrained high-order statistical blind deconvolution of spectral data," Chinese Optics Letters 3(9), 552-555 (2005)