基于高光谱成像的咸鸭蛋腌制品质快速检测

Rapid Detection on the Quality of Salted Duck Eggs Based on Hyperspectral Imaging

咸鸭蛋是一种中国传统腌制食品,水分和脂肪含量是评价咸鸭蛋腌制品质的重要指标。本文旨在采用高光谱成像系统结合化学计量学探究不同腌制阶段咸鸭蛋的水分和脂质含量变化及分布状态。实验采用高光谱成像技术获取432~961 nm波长范围内的咸鸭蛋剖面反射光谱。为了减小光谱信号的噪声,通过Savitzky-Golay平滑(SG)、高斯滤波(Gauss)和标准正态变换(SNV)三种方法对原始光谱进行预处理,使用竞争自适应加权采样算法(CARS)挑选最优波长,然后进一步采用偏最小二乘回归算法(PLSR)和人工神经网络(ANN)对水分和脂质含量进行定量预测。结果表明,ANN模型能够更好地对水分及脂质含量进行预测,对蛋白水分、蛋黄水分和蛋黄脂质的预测决定系数分别为0.9306,0.9552和0.8896。最后,为了更加直观、全面地对咸鸭蛋腌制过程中的水分和脂质含量进行评价,根据ANN模型绘制出咸鸭蛋剖面的水分及脂质含量分布图。可视化分布图成功显示了咸鸭蛋在不同腌制时间水分和脂质的空间分布。本研究为咸鸭蛋生产及下游食品加工企业提供了一种基于光谱技术的快速检测方法。

Salted duck eggs are a type of traditional Chinese pickled delicacy,and moisture and lipid content are important indexes for evaluating the quality during processing.This study used a hyperspectral imaging(HSI)system in conjunction with chemometrics to investigate the content change and distribution of moisture and lipid during different salting stages of duck eggs.The HSI was used to obtain reflectance spectral information of salted duck eggs in the 432~961 nm wavelength range.To minimize the noise in spectral signals,three preprocessing methods including Savitzky-Golay smoothing(SG),Gauss filter smoothing(Gauss),and standard normal variation(SNV)were used.The competitive adaptive reweighted sampling(CARS)was used to select the optimal wavelengths for predicting moisture and lipid content,and then the partial least squares regression(PLSR)and artificial neural network(ANN)methods were used to predict moisture and lipid content quantitatively.Results showed that ANN model could exhibited a better performance in predicting moisture and lipid content with coefficients of determination of the protein moisture,yolk moisture and yolk lipid of 0.9306,0.9552 and 0.8896 respectively.Finally,the ANN model was used to create a distribution map of moisture and lipid content in the profile of salted duck eggs.The visualization distribution maps successfully display the distribution of moisture and lipid content during different salting periods.This study could lead to the development of a rapid detection method based on spectrum technology for production and processing of salted duck egg industry.