应用独立分量分析提高动态光谱法的信噪比

Improving signal noise ratio of dynamic spectrum by ICA

为了提高动态光谱法的信噪比,针对动态光谱信号的特点和计算需求引入了独立分量分析(ICA)方法,基于负熵判据提出了ICA算法应用的具体实现步骤,进行了相关实验并对结果进行了讨论。结果证明,ICA算法可以在需要较少样本量的情况下有效降低动态光谱法中的噪声,相对于传统的相干平均方法,该方法可在提高信噪比的同时,提高动态光谱法的波长分辨率,为光谱数据的后期处理提供了可靠保证。

The Independent Component Analysis（ICA） is introduced in order to increase the Signal Noise Ratio（SNR） of Dynamic Spectrum（DS） which is a revolutionary method for blood component measuring.The neg-entrophy theory is introduced,serial of steps which are used in the data processing based on it are designed,and the corresponding experiments and discussion are done to test the method.Results show that through the ICA method,the noise can be decreased using less samples.Compared with the conventional superposition method,the ICA may increase the wavelength-resolution while the SNR increases which grantees the consequent processing of DS data.