Evaluation of quality of parboiled rice for export by near infrared spectroscopy

หมวดหมู่: On going work


The aim of this study will be to evaluate the quality of parboiled rice including classification of damaged paddy (i.e. red kernels, yellow kernels, chalkiness kernels, immature kernels and defected kernels), the texture of cooked parboiled rice and the relationship between chemical composition (i.e. amylose and lipid content) and texture of cooked parboiled rice. The paddy and parboiled rice samples will be collected from Capital Chainat Rice Mill company limited, in Saphaya District, Chainat Province, Thailand. In the case of classification, the damaged paddy will be used hyperspectral imaging (NIR) technique at the wavelength range 900-1700 nm and the develop the prediction model for classification of damaged paddy will be developed using partial least square discriminant analysis (PLS-DA). For the case of texture of cooked parboiled rice, FT-NIR spectrometer (MPA, Bruker, Germany) with wavelength 800-2500 nm and linear variable filter spectrometer (MicroNIR Pro, Viavi, USA) with wavelength from 950-1650 nm will be used. Each milled parboiled rice sample will be scanned in a quartz cup by scanning through the bottom of cup. The scanned sample is divided into 2 parts the first part is for determination of amylose and lipid content by Colorimetric methods and Acid hydrolysis solvent extraction using Soxtec™ method respectively, and the second part is for texture analysis. The relationship between chemical compostion (i.e. amylose and lipid content) and texture of parboiled rice will be established by two-way analysis of variance (ANOVA). The texture of cooked parboiled rice will be done by reference method by ISO 11747 Rice-Determination of Rice Kernel Resistance to Extrusion after Cooking method. The original and pre-treated spectral data and chemical data will be used to develop the predicting model for texture properties using partial least squares regression.

28 ก.ย. 2561

ผู้ชม 7 ครั้ง

Engine by shopup.com