光谱数据融合技术在食用菌质量评价中的应用

Application of Spectral Data Fusion Technology for Edible Fungi Quality Evaluation

光谱数据融合技术是可以将多种光谱设备数据进行信息融合来提高检测结果准确度的一种新兴的快速检测手段。该技术运用化学计量学方法,从优化光谱数据、特征选择2个方面对数据进行预处理,以消除无关变量的干扰,有效提取信息。采用低级、中级、高级数据融合技术建立最终分类或预测模型,运用该模型进行检测可以弥补单一检测设备的不足,具有广阔的应用前景。本文结合光谱数据融合技术在国内外的最新研究进展,对该技术在食用菌的产地溯源、储存年限、种类鉴别、理化指标含量预测等方面的应用进行综述,为完善食用菌质量评价体系提供参考。

Spectral data fusion is a new and speedy detection technology which can merge a variety of spectral equipment data to improve the accuracy of experimental results.This technique uses the chemometric method to preprocess output data from two aspects:optimization of spectral data and feature selection,aiming to eliminate the interference of irrelevant variables and to extract the information effectively as well.Adopting the fusion technology of those low-level,intermediate-level and advanced data for conducting final classification or making a prediction model as well as operating the model to compensate for the shortcomings of a single equipment detection should have a broad prospect of application.This paper summarizes the current application of this technology for quality evaluation of edible fungi,such as origin tracking,storage life,species identification,prediction of physicochemical index or content and so on,with a purpose to provide references for future improvement of the quality evaluation system in the edible fungi.