Research Of Improvement Technology Of Spectrum Resolution Based On Nonlinear Coordinate Transform Combined With Deconvolution Algorithm

As an important means of nondestructive material analysis, spectral analysis has been widely used in many fields such as remote sensing, exploration, biological detection.The emission and reflectance spectral characteristics of the material can parse the material’s composition. But the spectral characteristics of the dataobtainedby spectrometers not only contains the characteristics of the target, at the same time introduces the responsecharacteristics ofthe equipment themselves.Spectral response characteristics can be measured by spectral resolution, namely the ability that determines and separates thespectrum characteristic peaks.The higher spectral resolution of spectrometer, the smaller the impact of the target spectrum with the device itself, but usually the response and SNR performance will be limited. On the other hand, the response function of the spectrometer will cause the target spectral bandwidth increase and spectral overlap phenomenon, which can lead to the quality of the measured spectral data declined. So the process of the spectral acquisition is a convolution process.In an attempt to enhance the resolution of the spectral data, wecan use the deconvolution algorithm, the instrument response function as a convolution kernels, to decrease the instrument features’ s influence of the target spectral characteristics.With acousto-optic tunable filter(AOTF) as the core of spectral components of the spectrometer, the spectral response has the characteristics of spectral bandwidth changes with diffraction wavelength, namely different wavelengths corresponding to different spectral resolution, which can cause in conventional deconvolution algorithm, the convolution kernel in different diffraction wavelength position is different.So, this article adopts a method that is nonlinear scale transform, which the wavelength and spectral resolution will be equivalently transformed.Under the condition of energy consistent, spectrum dimension coordinates will be stretched.After the transformation, the spectral resolution be consistent in spectral dimension, so that solving the deconvolutionproblems.Through the inverse transform back to the original coordinate system, the limitation problem of the deconvolution algorithm will be solved.Firstly, simulation under the AOTF spectrometer spectral resolution characteristics, the impulse signal, the response of the twin peaks overlapping signals, multispectral overlapping signal, usethe methods of FSD and nonlinear iterative deconvolution respectively combining with the nonlinear coordinate transform toenhance the spectral resolution.By comparison with the simulation results, the method of the nonlinear iterative deconvolution combining with the nonlinear coordinate transform is much better than FSD combining with the nonlinear coordinate transform in spectral resolution enhancing.Baseline separation of overlapped peaks can be achieved, spectral absorption depth increased, the spectral characteristics of the more obvious, more close to real spectral curve features.Secondly, using the laboratory existing AOTF test platform, to get NeAr the standard lamp and the response of the atmospheric transmission spectra, using nonlinear coordinate transformation and nonlinear iterative deconvolution combination of methods to improve the resolution the measured data.Finally, using the laboratory existing AOTF test platform, to get NeArcalibrationlamp.Results show that the target spectrum half tall wide average 48.9% reduction, spectral absorption depth can increase 25 times, peak to valley ratio improved.Thirdly, the response of the atmospheric transmission spectra, using nonlinear coordinate transformation and nonlinear iterative deconvolution combination of methods to improve the resolution the measured data. Atmosphere after wide spectrum data processing results show that in 950 nm, 970 nm, 1130 nm, 1430 nm, 1790 nm, 1960 nm, 2500 nm absorption characteristics, is appeared after deconvolution processing, and PcModwin of atmospheric real simulation comparison of the spectrum, these absorption characteristics are mainly water, CO2, and thus the above method has been effectively.