水果缺陷和内部品质同时在线检测方法研究

发布时间：2018-12-06 21:07

【摘要】：水果缺陷和内部品质同时在线检测技术具有重要现实意义,可以发现水果的内外部缺陷并及时剔除,提高了水果的内部品质,实现水果优质高价的同时提高了市场竞争力。在对外贸易中,有利于减少果农在出口创汇中的损失,提高水果产业附加值水平。与传统的有损检测相比,可见/近红外光谱漫透射技术不用破损样品,操作简便,具在线检测技术优势,在很大的程度上提高了检测的速度,准确性。本文以鸭梨黑心病,大黄桃表面缺陷和柑橘浮皮果为研究内容,探究同时在线检测水果缺陷和内部品质的可行性。本研究为打破技术垄断,降低外贸退货率,为果农增收创汇,提供了科学的参考和有力的技术支持。本研究的结果和结论如下所述:(1)采用可见/近红外光谱技术在线检测鸭梨黑心病,建立了峰面积判别法,主成分分析法(PCA)和偏最小二乘判别法(PLS-DA)的定性判别模型。其中PLS-DA判别模型对鸭梨黑心病的判别正确率为100%。同时,建立了鸭梨剔除黑心病梨的正常果内部品质的偏最小二乘模型,最优模型的Rp为0.93,RMSEP为0.45。Brix。(2)采用可见/近红外光谱技术在线检测大黄桃表面缺陷,建立了主成分分析法(PCA),最小二乘支持向量机(LS-SVM)和偏最小二乘判别法(PLS-DA)的定性判别模型。其中,PLS-DA判别模型对大黄桃表面缺陷的判别正确率为100%。同时,建立了大黄桃剔除表面缺陷果的正常果内部品质的偏最小二乘模型,最优模型的Rp为0.95,RMSEP为0.71。Brix。(3)采用可见/近红外光谱技术在线检测柑橘浮皮果,建立了主成分分析法(PCA),反向传播神经网络(BP-ANN),最小二乘支持向量机(LS-SVM)和偏最小二乘判别法(PLS-DA)的判别正确率为100%。同时,建立了柑橘剔除浮皮果的正常果内部品质的偏最小二乘模型,最优模型的Rp为0.88,RMSEP为0.77。Brix。(4)采用可见/近红外漫透射技术结合化学计量学方法,进行了水果缺陷和内部品质的同时在线检测,结果显示检测效率在96%以上。研究结果表明,同时在线检测水果缺陷和内部品质是可行的。
[Abstract]:On-line detection of fruit defects and internal quality has important practical significance. It can find the internal and external defects of fruit and eliminate them in time, improve the internal quality of fruit, realize the high price of fruit quality and improve the market competitiveness. In foreign trade, it helps to reduce the loss of fruit farmers in export and increase the added value of fruit industry. Compared with traditional lossy detection, the technique of diffuse transmission of visible / near infrared spectroscopy is easy to operate and has the advantage of on-line detection, and improves the speed and accuracy of detection to a great extent. In this paper, the black heart disease of Yali pear, the surface defect of large yellow peach and the citrus flowered fruit were used as the research contents to explore the feasibility of on-line detecting the defects and internal quality of the fruit at the same time. This study provides scientific reference and strong technical support for breaking technology monopoly, reducing foreign trade return rate, increasing fruit farmers' income and earning foreign exchange. The results and conclusions of this study are as follows: (1) using visible / near infrared spectroscopy (NIR) to detect the black heart disease of Yali pear on line, a peak area discrimination method is established. The qualitative discriminant model of principal component analysis (PCA) and partial least squares discriminant (PLS-DA). The PLS-DA discriminant model was used to distinguish the black heart disease of Yali pear. At the same time, a partial least square model was established for removing the internal quality of black heart pear from Yali pear. The Rp of the optimal model was 0.93 Rp = 0.45.Brix.2.The visible / near infrared spectroscopy was used to on-line detect the surface defects of large yellow peach. The qualitative discriminant models of (PCA), least squares support vector machine (LS-SVM) and partial least squares discriminant method (PLS-DA) based on principal component analysis (PCA) are established. The PLS-DA discriminant model was used to distinguish the surface defects of yellow peach. At the same time, a partial least square model of normal fruit internal quality was established for removing surface defects of large yellow peach. The Rp of the optimal model was 0.95 Rp = 0.71.Brix. (3) the visible / near infrared spectroscopy was used to on-line detect the floating peel of citrus. In this paper, (PCA), back propagation neural network (BP-ANN), least square support vector machine (LS-SVM) and partial least squares discriminant method (PLS-DA) are established. At the same time, a partial least square model for the internal quality of the normal citrus fruit was established. The Rp of the optimal model was 0.88C RMSEP was 0.77.Brix. (4) the visible / near infrared diffuse transmission technique was used in combination with chemometrics. On-line detection of fruit defects and internal quality was carried out, and the results showed that the detection efficiency was over 96%. The results show that it is feasible to detect fruit defects and internal quality simultaneously.
【学位授予单位】：华东交通大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：S66;TP274

【相似文献】