Non‐destructive technique for determining the viability of soybean (Glycine max) seeds using FT‐NIR spectroscopy. (17th October 2017)

• Record Type: Journal Article
• Authors: Kusumaningrum, Dewi; Lee, Hoonsoo; Lohumi, Santosh; Mo, Changyeun; Kim, Moon S; Cho, Byoung‐Kwan
• Is Part Of: Journal of the science of food and agriculture. Volume 98:Number 5(2018)
• Page Start: 1734
• Page End: 1742
• Publication Date: 2017-10-17
• Subjects: soybean -- viability -- non‐destructive measurement -- near infrared spectroscopy; Food -- Periodicals; Agriculture -- Periodicals; 664
• DOI: 10.1002/jsfa.8646
• Languages: English
• ISSNs: 0022-5142

Abstract

Abstract: BACKGROUND: The viability of seeds is important for determining their quality. A high‐quality seed is one that has a high capability of germination that is necessary to ensure high productivity. Hence, developing technology for the detection of seed viability is a high priority in agriculture. Fourier transform near‐infrared (FT‐NIR) spectroscopy is one of the most popular devices among other vibrational spectroscopies. This study aims to use FT‐NIR spectroscopy to determine the viability of soybean seeds. RESULTS: Viable and artificial ageing seeds as non‐viable soybeans were used in this research. The FT‐NIR spectra of soybean seeds were collected and analysed using a partial least‐squares discriminant analysis (PLS‐DA) to classify viable and non‐viable soybean seeds. Moreover, the variable importance in projection (VIP) method for variable selection combined with the PLS‐DA was employed. The most effective wavelengths were selected by the VIP method, which selected 146 optimal variables from the full set of 1557 variables. CONCLUSIONS: The results demonstrated that the FT‐NIR spectral analysis with the PLS‐DA method that uses all variables or the selected variables showed good performance based on the high value of prediction accuracy for soybean viability with an accuracy close to 100%. Hence, FT‐NIR techniques with a chemometric analysis have the potential for rapidly measuring soybean seed viability. © 2017 Society of Chemical Industry