基于相关分析技术的近红外光谱信息特征提取

运用信号分析的理论，研究了把相关分析技术应用于近红外光谱信息特征提取的方法.结果表明：通过相关处理，光谱信息特征得到分离和增强，可以作为一种光谱信息特征提取的有效方法用于定性和定量分析，为解决近红外光谱检测中光谱叠加、信息率低等难题提供了一种方法。

Study on Feature Extraction of Near-infrared Spectroscopy by Correlation Analysis Technique

From the point of view of signal analysis, a new way to feature extraction of near infrared reflection spectroscopy(NIRDRS) by the correlation analysis theory was discussed. Taking instant food condiment as object, this paper analyzed the coherent of spectrum between pure components and between pure component and different component content mixture. Through correlated processing, the information feature of spectroscopy could be separated and strengthened to be used for qualitatively and quantitatively spectrochemical analysis. The results showed that it was an effective method in spectrum analysis, and also a new resolution to the problems of spectrum superposition and low information rate in NIRDRS detection.