小波变换应用于红外光谱法测定西红柿中葡萄糖含量

Application of Wavelet Transform in IR-Spectrometric Determination of Glucose in Tomato

在红外光谱法测定西红柿中葡萄糖含量中,应用小波变换处理技术使西红柿样品的红外光谱图去噪声和平滑化,从而增强了谱图的分辨能力。选择在2 900 cm^(-1)附近由醛基所引起的吸收峰作为特征峰,分别建立了由葡萄糖标准的质量值(mg)作为自变量m和对应的吸光度作为因变量A作图而得到的线性和非线性数学模型,并应用于定量分析。非线性回归方程的相关性较好,其相关系数达0.938 3,而线性回归方程的相关系数仅达0.801 5,而且在分析中应用非线性数学模型时,所得测定结果的准确度相对较高,其相对误差为—2.29%。

Wavelet transform method was applied to the IR-spectrometric determination of glucose in tomato to de-noise and smoothen the IR spectrum of tomato sample, resulting enhancement of resolution of the spectrum. In the determination, wave numbers close to 2 900 cm^-1 （due to-CHO group） were chosen as the characteristic absorption peaks. By plotting values of mass （rng） of standard glucose （as independent variables m） against values of the respective absorhance （as dependent variables A）, both the linear and non-linear mathematical models were established for the quantification. It was shown that the correlation coefficient of non-linear regression equation attained to a value of 0. 938 3, while a value of only 0. 801 5 was attained for the linear regression equation, and that results with higher accuracy were obtained by using the non-linear mathematical model, giving relative error of --2. 29%.