激光与光电子学进展, 2020, 57 (1): 013002, 网络出版: 2020-01-03   

基于近红外光谱检测不同产地石榴的糖度 下载: 1213次

Detection of Sugar Content of Pomegranates from Different Producing Areas Based on Near-Infrared Spectroscopy
作者单位
华东交通大学机电与车辆工程学院水果智能光电检测技术与装备国家地方联合工程研究中心, 江西 南昌 330013
摘要
基于可见/近红外漫透射光谱技术,探讨了对不同产地石榴的品质进行快速无损检测的可行性。首先,利用能够有效抑制杂散光影响的动态在线检测装置采集石榴的近红外光谱,测试石榴糖度的真值;然后结合主成分分析法对不同产地的石榴进行聚类分析,大致将样品分为两类;最后建立不同产地石榴的偏最小二乘判别分析模型,模型的判别准确率为97%以上。采用多种预处理方法(S-G平滑、归一化、基线校正、MSC等)建立了两类石榴的单一模型,结果表明:基线校正的效果明显优于其他方法,所建立的四川石榴模型的预测集相关系数Rp为0.82,预测集均方根误差(RMSEP)为0.37,建模集相关系数Rc为0.90,建模集均方根误差(RMSEC)为0.31;云南石榴模型的Rp为0.81,RMSEP为0.33,Rc为0.87,RMSEC为0.27。在后期采用未参与建模的样品的分选验证实验中,两个产地石榴的判别率为95%,糖度的分选准确率可达92.5%。结果表明,近红外光谱在石榴产地的判别和糖度的分选上具有重要意义,可为以后的石榴在线分选研究提供依据。
Abstract
In this study, the feasibility of the rapid non-destructive testing method for the pomegranate quality in the Sichuan and Yunnan Provinces is investigated based on the visible/near-infrared diffuse transmission spectroscopy technique. First, the near-infrared spectra of pomegranates are obtained using a dynamic online detection device which can effectively suppress the effect of stray light, and the actual sugar content value is measured. In combination with the principal component analysis method, the cluster analysis of pomegranates from different producing areas can approximately divide the samples into two categories. Further, a partial least squares discrimination analysis model is developed for the pomegranates from two distinct producing areas, which exhibits an accuracy greater than 97%. Meanwhile, multiple pretreatment methods, such as Savitzky-Golay smoothing, normalization, baseline correction, and multiplicative signal correction, are employed to establish a single model for two pomegranate types. Based on the obtained results, the baseline correction method is observed to be better than the other examined methods. In particular, the correlation coefficient of the prediction set (Rp) of the established Sichuan pomegranate model is 0.82, the root mean square error of prediction set (RMSEP) is 0.37, the correlation coefficient of the calibration set (Rc) is 0.90, and the root mean square error of calibration set (RMSEC) is 0.31. However, for the Yunnan pomegranate model, the Rp is 0.81, the RMSEP is 0.33, the Rc is 0.87, and the RMSEC is 0.27. In the post-sorting verification experiment for samples not involved in modeling, the discriminating rate of pomegranates in both the producing areas is 95%, whereas the sugar content sorting accuracy is 92.5%. Thus, the near-infrared spectroscopy is of considerable significance with respect to the discrimination of the pomegranate producing area and the sorting of its sugar content and may form the basis for future pomegranate online sorting research.

刘燕德, 张雨, 徐海, 姜小刚, 王军政. 基于近红外光谱检测不同产地石榴的糖度[J]. 激光与光电子学进展, 2020, 57(1): 013002. Yande Liu, Yu Zhang, Hai Xu, Xiaogang Jiang, Junzheng Wang. Detection of Sugar Content of Pomegranates from Different Producing Areas Based on Near-Infrared Spectroscopy[J]. Laser & Optoelectronics Progress, 2020, 57(1): 013002.

本文已被 6 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!