摘要
To subtract the slit function from the measured spectrum, a wavelet-based deconvolution method is proposed to obtain a regularized solution of the problem. The method includes reconstructing the signal from the wavelet modulus maxima. For the purpose of maxima selection, the spatially selective noise filtration technique was used to distinguish modulus maxima produced by signal from the one created by noise. To test the method, sodium spectrum measured at a wide slit was deconvolved. He-Ne spectrum measured at the corresponding slit width was used as slit function. Sodium measured at a narrow slit was used as the reference spectrum. The deconvolutton result shows that this method can enhance the resolution of the degraded spectrum greatly.
为消除测量光谱中狭缝宽度的影响,提出了一种基于小波的反卷积算法以获得问题的规整化解,该方法使用一维多尺度边缘重构技术.空域选择滤波算法被用来区分由信号产生的极值与由噪声产生的极值.实验中,对宽狭缝时测得的Na黄光2条特征谱线进行反卷积处理,以相应狭缝宽度时测得的He-Ne光谱作为狭缝函数,并与较窄狭缝时测得的Na光谱作了比较.反卷积结果显示本方法可以极大地提高测量光谱的分辨率.
