文章摘要
陈李品;于繁千惠;陶然;陈桂东;李兆杰;薛长湖.基于高光谱成像技术预测牡蛎干制加工过程中的水分含量[J].中国食品学报,2020,20(7):261-268
基于高光谱成像技术预测牡蛎干制加工过程中的水分含量
Prediction of Moisture Content in Oyster Drying Process Based on Hyperspectral Imaging
  
DOI:
中文关键词: 牡蛎  水分  高光谱成像技术  化学计量学
英文关键词: oysters  moisture content  hyperspectral imaging technique  chemometrics
基金项目:国家自然科学基金面上项目(31871867);国家贝类产业技术体系项目(CARS-49)
作者单位
陈李品;于繁千惠;陶然;陈桂东;李兆杰;薛长湖 中国海洋大学食品科学与工程学院 
摘要点击次数: 9
全文下载次数: 0
中文摘要:
      提出一种应用高光谱成像技术结合化学计量学检测牡蛎干制加工过程中水分含量的方法。采用高光谱成像系统,在400~1 100 nm范围内,采集到5个干燥时期的100个牡蛎干样本高光谱图像。提取所有样本感兴趣区域的平均光谱数据,对原始光谱数据进行多元散射校正(MSC)、卷积平滑(S-G)预处理,采用相关系数法提取8个特征波长。基于所提取的特征波长,建立光谱数据与水分含量的多元线形回归(MLR)和BP神经网络模型。结果表明:两种模型均有较好的预测效果。MLR模型的校正集、预测集和交叉验证集的相关系数较BP神经网络低;校正集、预测集和交叉验证集均方根误差分析结果表明,BP神经网络效果较MLR好。高光谱成像技术结合化学计量学方法可检测牡蛎干制过程中水分含量的变化。
英文摘要:
      The suitability of hyperspectral imaging technology combined with chemometrics was investigated to determine the moisture content of dried oysters. The moisture content not only affects the storage of scallops, but also is closely related to the quality characteristics such as texture and texture. In this experiment, a hyperspectral imaging system was used to collect hyperspectral images in the range of 400 to 1 100 nm. A total of 100 oysters samples of hyperspectral images in 5 different drying stages were collected. The average spectral data of all the regions of interest of the sample are extracted, and the original spectral data is subjected to multi-dimensional scatter correction (MSC) and convolution smoothing (S-G) for preprocessing, and eight characteristic wavelengths are extracted by a correlation coefficient method. Based on the extracted characteristic wavelengths, multivariate linear regression(MLR) and BP neural network models of spectral data and moisture content were established. The results show that both models have achieved good prediction results. The correlation coefficient between the correction set, prediction set and cross-validation set of MLR model are lower than that of BP neural network, but from the analysis of RMSEC, RMSEP and RMSECV, the effect of BP neural network are better than that of MLR. The results show that hyperspectral imaging techniques combined with chemometric methods can be used to detect the moisture content of oysters during drying.
查看全文   查看/发表评论  下载PDF阅读器
关闭

手机扫一扫看
分享按钮