A Novel Spectrum Profile Fitting Method of Non-Gaussian Model

摘要

Spectrum extraction plays a crucial role in two-dimensional spectrum data processing. One of the most important spectrum extraction methods is profile fitting algorithm. In this article we studied the spectrum profile fitting method of non-Gaussian model. The general solution methods always have poor adaptability and some limitation, in order to approximate spectral profiles better, we introduced a multi-Gauss-Lorenz curve superimposed model, then the Nelder-Mead simplex iterative method is used to optimize the objective function, so that the method gains an adaptability for various profile shapes. Experiments are carried out on LAMOST data to demonstrate the effectiveness of this algorithm. It has been proved that this proposed method has a good performance.