Wavelet Transform Zero-Crossing Method for Auto-Extraction of Spectral Lines
ZHAO Rui-zhen1, HU Zhan-yi1,ZHAO Yong-heng2
1. National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing 100080, China 2. National Astronomical Observatory, Chinese Academy of Sciences, Beijing 100012, China
Abstract：Wavelet transform method is used in this paper. The authors firstly studied the properties of spectral lines in the wavelet domain. By introducing the items of up-zero-crossings and down-zero-crossings, the authors concluded that absorption lines and emission lines correspond to different kinds of zero-crossings. Therefore the authors presented a new wavelet transform zero-crossing method for the extraction of spectral lines and the normalization of continuum spectra. The useful spectral lines can be obtained together with the continuum spectra. The extraction of spectral lines in this paper is a direct method, which avoids the errors caused during the process of noise reduction. The experiments both on stars and nearby galaxies show that our method can be used to accurately extract the spectral lines, which is helpful to the computation of characteristic parameters and the automatic classification of spectra based on spectral lines.
Key words：Wavelet transform;Astronomical spectra;Spectral line extraction;Zero-crossings
