Browsing by Author "Giraldo Ossa, Erika Fernanda"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
Item Calibración en harina de arroz blanco para la determinación de las variables de SB y BD del perfil de viscosidad por la técnica espectroscopía de reflectancia en el infrarrojo cercano (NIRS)(Universidad Santiago de Cali, 2023) Giraldo Ossa, Erika Fernanda; Orduña Ortega, Julieth (Directora); Loaiza de la Pava, Jhoana Katerine (Codirectora)Currently, the importance has increased for the genetic improvement program of the Latin American Irrigated Rice Fund (FLAR) to evaluate the quality of rice through quick, economical, environmentally friendly tools that require a minimum of preparation for your analysis. Given the above, the possibility of using the NIRS near-infrared spectroscopy technique (400-250nm) was studied through the development of a calibration equation to measure the Breakdown (BD) and Setback (SB) variables, which currently are evaluated in the Rapid Visco Analyzer RVA equipment (by its name in English), these viscosity variables of rice starch allow us to associate the cooking quality of the grain. The implementation of this technique seeks to facilitate the analysis of early generation rice lines, since currently, due to the capabilities of the RVA, it is only possible to evaluate advanced lines. The calibration was carried out on a set of 196 samples with a mathematical treatment “2,16,32,1”, an MPLS partial least squares regression model and SNV and Detrend smoothing, where the following statistical data of the variables were obtained. viscosity of BD and SB which were RSQ calibration (0.864 and 0.853), SEC (129.580 and 234.552), 1-VR (0.786 and 0.761), SECV (350.939 and 635.427) and RPD (2.168 and 2.049) respectively. Subsequently, the developed equation was verified through internal and external validation, obtaining the following internal RSQ statistics (0.822 and 0.808) and a SEP (151.909 and 309.770), RSQ external (0.791 and 0.531) and a SEP (210.444 and 944.374) respectively for the BD and SB variables. The calibration equation presents a good prediction for routine evaluation in early generation samples for the BD variable.