Authenticity Discrimination and Adulteration Level Detection of Camellia Seed Oil via Hyperspectral Imaging Technology

Yuqian Shang,Liwei Bao,Haiwen Bi,Shihao Guan,Jiafeng Xu,Yuqi Gu,Chao Zhao

DOI: https://doi.org/10.1007/s12161-024-02577-x

IF: 2.9

2024-01-26

Food Analytical Methods

Abstract:Camellia seed oil has high nutritional value with expensive price, and it is often adulterated other low-price vegetable oil to reduce cost. To identify the authenticity and adulteration level of camellia seed oil, a rapid nondestructive detection method based on hyperspectral imaging technology was proposed in this study. Binary adulterated samples were obtained. Camellia seed oil was adulterated soybean oil by the concentration of 0~100%, and the corresponding hyperspectral data of these samples were collected. Two qualitative discrimination methods including support vector machine (SVM) and linear discriminant analysis (LDA) were employed to qualitative discriminate the authenticity of the camellia seed oil. Meanwhile, two quantitative prediction models were established by partial least squares regression (PLS) and synergy interval partial least squares regression (Si-PLS) to predict the adulterated level of Camellia seed oil. The results showed that the accuracy of SVM and LDA models established after pretreatment to discriminate the authenticity of camellia seed oil could reach 100%. For the quantitative prediction of the adulterated level of camellia seed oil, the optional prediction model was developed from Si-PLS after the hyperspectra were continuously preprocessed by second derivative (2-Der) and standard normal variate (SNV) transform. The correlation coefficient of the prediction set (R 2 P ) and the root mean square error of prediction set (RMSEP) of the Si-PLS model were 0.9789 and 4.69%, respectively. Finally, the visual prediction images with different adulteration level of camellia seed oil were realized using the optimal Si-PLS prediction model. As the consequence, the hyperspectral imaging technology was a potential method for detecting the authenticity and adulteration level of camellia seed oil.

food science & technology