Hyperspectral imaging and partial least square discriminant analysis for geographical origin discrimination of white rice

Authors: MO, CHANGYEUN - Rural Development Administration - Korea; LIM, JONGKUK - Rural Development Administration - Korea; KWON, SUNG - Seoul National University; LIM, DON - Seoul National University; Kim, Moon; KIM, GIYOUNG - Rural Development Administration - Korea; KANG, HUNGSOOK - Rural Development Administration - Korea; KWON, KYUNG-DO - Rural Development Administration - Korea; CHO, BYOUNG-KWAN - Chungnam National University

Submitted to: Journal of Biosystems Engineering
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/8/2017
Publication Date: 12/15/2018
Citation: Mo, C., Lim, J., Kwon, S., Lim, D., Kim, M.S., Kim, G., Kang, H., Kwon, K., Cho, B. 2018. Hyperspectral imaging and partial least square discriminant analysis for geographical origin discrimination of white rice. Journal of Biosystems Engineering. https://doi.org/10.5307/JBE.2017.42.4.293.
DOI: https://doi.org/10.5307/JBE.2017.42.4.293

Interpretive Summary: Identifying the geographical origin of rice is important when an economic motivation to misrepresent origin on product labels may exist due to pricing differences or due to consumer resistance to specific origins as a result of concern about quality or safety. Awareness of potentially mislabeled products can increase consumer distrust and create economic loss for producers whose products are accurately labeled. DNA analysis methods can discriminate between different rice cultivars, but require expert skill and are not useful for comparing samples of the same cultivar grown in different regions. However, some laboratory analysis methods to determine chemical composition have been found useful for identifying different geographical origins of same-cultivar rice samples. This study sought to develop a rapid method for discriminating between rice samples of different geographical origin using visible and near-infrared (VNIR) hyperspectral reflectance imaging of the samples, since VNIR spectral data are based on chemical composition. Discrimination model development included investigation of spectral characteristics of South Korean and Chinese white rice samples, spectral pretreatments, and effective spatial imaging resolution. The results demonstrated over 99% accuracy in discriminating between white rice samples from South Korea and China. This research will benefit those seeking to develop and implement methods for rapidly identifying rice origin for product verification, to protect rice producers and increase consumer trust.

Technical Abstract: Purpose: This study aims to propose a method for fast geographical origin discrimination between domestic and imported rice using a visible/near-infrared (VNIR) hyperspectral imaging technique. Methods: Hyperspectral reflectance images of South Korean and Chinese rice samples were obtained in the range of 400 nm to 1000 nm. Partial least square discriminant analysis (PLS-DA) models were developed and applied to the acquired images to determine the geographical origin of the rice samples. Results: The optimal pixel dimensions and spectral pretreatment conditions for the hyperspectral images were identified to improve the discrimination accuracy. The results revealed that the highest accuracy was achieved when the hyperspectral image’s pixel dimension was 3.0 mm × 3.0 mm. Furthermore, the geographical origin discrimination models achieved a discrimination accuracy of over 99.99% upon application of a first-order derivative, second-order derivative, maximum normalization, or baseline pretreatment. Conclusions: The results demonstrated that the VNIR hyperspectral imaging technique can be used to discriminate geographical origins of rice.