论文部分内容阅读
提出一种光谱信号噪声的乘性加性混合分析模型,并采用维纳滤波和同态滤波相结合的算法对光谱信号进行去噪处理。仿真结果表明,该算法比移动平均算法、最小均方算法和递归最小均方算法具有更好的去噪性能。实验结果表明,氙灯光谱信号中的噪声符合乘性加性混合模型。与移动平均算法、最小均方算法和递归最小均方算法相比,从该算法处理后的汞灯光谱信号中能够提取更加稳定的谱峰谷位置、谱峰幅度、谱峰半峰全宽等特征值,定量分析时能获得更好的结果。
A hybrid model of spectral signal noise multiplication and additive analysis is proposed. The spectral signal is denoised by a combination of Wiener filter and homomorphic filter. Simulation results show that the proposed algorithm has better denoising performance than the moving average algorithm, the least mean square algorithm and the recursive least mean square algorithm. The experimental results show that the noise in the xenon lamp spectral signal conforms to the multiplicative additive model. Compared with the moving average algorithm, the least mean square algorithm and the recursive least mean square algorithm, the more stable spectrum peak position, the peak amplitude, the full width at half maximum of the spectrum peak, etc. can be extracted from the processed mercury lamp spectral signal Eigenvalues, quantitative analysis can get better results.