小波去噪与平均算法去噪在拉曼光谱中的应用

Application of adaptive wavelet threshold denoising and average arithmetic in Raman spectra

拉曼光谱分析中,噪声的存在常影响分析的准确度和检测限,现有滤波方法在光谱信号除噪方面有种种缺陷。使用自适应小波阈值函数滤噪法和平移不变量小波去噪法两种方法,并分别与传统方法平均算法相结合,实现了信号与噪音的有效分离,均取得了很好的效果。即使对信噪比小于1的高噪声信号也能够很好地保留了信号的细节,获取满意的处理结果。

During the spectrum analysis process, the noise usually influences the analytical accuracy and the detection limit. The existing filtering methods have various flaws in filtering the noise of spectral signal. Two methods are described and employed. The first one is adaptive wavelet threshold function denoising method, and another is a denoising method of wavelet transform based on translation invariance. At the same time, they are integrated with the classical average arithmetic respectively, then the signal is separated from the noise effectively and the result is satisfactory by using these integrated methods. These methods also can keep the main edges of the signal and obtain satisfactory processing results even for the signal whose signal-to-noise ratio is less than one.