基于近红外光谱的大黄鱼新鲜度评价模型

目的 探索定量评价大黄鱼新鲜度的方法。方法 在整鱼背部采集近红外光谱, 将原始光谱预处理后分别与挥发性盐基氮(TVB-N)、菌落总数建立偏最小二乘(PLS)模型、区间偏最小二乘(iPLS)模型、向后区间偏最小二乘(biPLS)模型和联合区间偏最小二乘(siPLS)模型。结果 biPLS模型的精度最高、预测性能最佳。TVB-N的biPLS模型的校正集和预测集相关系数分别为0.8371和0.7652; 菌落总数的biPLS模型的校正集和预测集相关系数分别为0.878和0.7009。结论 大黄鱼的近红外光谱信息与其TVB-N、菌落总数间都存在较高的相关性, 所建模型可以快速、无损地定量评价大黄鱼的新鲜度。

Freshness evaluation model of Pseudosciaena crocea based on near-infrared spectra

Objective To investigate a method for the quantitatively freshness evaluation of Pseudosciaena crocea. Methods Near-infrared spectra of the whole back of fish was adopted and preprocessed. Quantitative models of total volatile basic nitrogen (TVB-N) content and aerobic plate count were built with the processed spectra, respectively. The partial least squares (PLS), interval PLS (iPLS), backward interval partial least squares (biPLS) and synergy interval partial least squares (siPLS) algorithms were used for modeling. Results biPLS model had the highest accuracy and predicted the best performance. The optimal biPLS model of TVB-N was achieved with correlation coefficient (Rc=0.8371) in calibration set and correlation coefficient (Rp=0.7652) in prediction set. The optimal biPLS model of aerobic plate count was achieved with correlation coefficient (Rc=0.878) in calibration set and correlation coefficient (Rp=0.7009) in prediction set. Conclusion There is a high correlation between near-infrared spectra and TVB-N or aerobic plate count. Near-infrared spectroscopy with biPLS can be successfully applied as an accurate and non-destructive method for the determination of freshness of Pseudosciaena crocea.