Title: The application of wavelet transform in processing the signal of solar spectrograph
Abstract:The solar spectrograph firstly takes the real-time accurate tracking of sun, then detects all monochromatic light within a certain wavelength simultaneously by making use of grating splitter, and fina...The solar spectrograph firstly takes the real-time accurate tracking of sun, then detects all monochromatic light within a certain wavelength simultaneously by making use of grating splitter, and finally obtains the continuous atmospheric parameter with high precision. In order to improve the accuracy of data and simplify the structure of the optical system, this article introduces the application of wavelet transform in removing noise and correcting baseline for the data measured. Baseline wander is mainly concentrated in the wavelet coefficients of low frequency band, while useful spectrum signal usually distributes in the wavelet coefficients of high frequency, and baseline wander can be removed by setting approximation coefficients to zero. Because of the time-frequency locality of wavelet transform, on the same frequency band of wavelet domain, the peak of the spectrum is compressed in few time channels. As a result, the amplitude of corresponding wavelet coefficients is bigger; On the contrary, the wavelet coefficients of noise is distributed on all the time channels with smaller amplitude of wavelet coefficients, so that the threshold method is used in this article to remove noise. The experiment shows that it is more suitable to decompose the spectrum signal at level 15 on the wavelet basis of sym5 with the approximation coefficients being set to zero and to process the detail coefficients by adaptive hard threshold. Examples show that this method provides an effective way to remove noise and correct baseline in the solar spectrum.Read More
Publication Year: 2017
Publication Date: 2017-10-24
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot