Memorias de investigación
Artículos en revistas:
Comparison of different pre-treatments to improve accuracy of total soluble solids content prediction models in grapes using a portable NIR spectrophotometer
Año:2014
Áreas de investigación
  • Agricultura
Datos
Descripción
The applicability of a portable NIR spectrometer for estimating the °Brix content of grapes by non-destructive measurement has been analysed in field. The NIR spectrometer AOTF-NIR Luminar 5030, from Brimrose, was used. The spectrometer worked with a spectral range from 1100 to 2300 nm. A total of 600 samples of Cabernet Sauvignon grapes, belonging to two vintages, were measured in a non-destructive way. The specific objective of this research is to analyse the influence of the statistical treatment of the spectra information in the development of °Brix estimation models. Different data pretreatments have been tested before applying multivariate analysis techniques to generate estimation models. The calibration using PLS regression applied to spectra data pretreated with the MSC method (multiplicative scatter correction) has been the procedure with better results. Considering the models developed with data corresponding to the first campaign, errors near to 1.35 °Brix for calibration (SEC = 1.36) and, about 1.50 °Brix for validation (SECV = 1.52) were obtained. The coefficients of determination were R2 = 0.78 for the calibration, and R2 = 0.77 for the validation. In addition, the great variability in the data of the °Brix content for the tested plots was analysed. The variation of °Brix on the plots was up to 4 °Brix, for all varieties. This deviation was always superior to the calculated errors in the generated models. Therefore, the generated models can be considered to be valid for its application in field. Models were validated with data corresponding to the second campaign. In this sense, the validation results were worse than those obtained in the first campaign. It is possible to conclude in the need to realize an adjustment of the spectrometer for each season, and to develop specific predictive models for every vineyard
Internacional
Si
JCR del ISI
No
Título de la revista
Journal of Food, Agriculture and Environment
ISSN
1459-0263
Factor de impacto JCR
0,44
Información de impacto
Volumen
12 (2)
DOI
Número de revista
Desde la página
218
Hasta la página
223
Mes
ABRIL
Ranking
Esta actividad pertenece a memorias de investigación
Participantes
  • Autor: Silvia Arazuri Universidad Pública de Navarra
  • Autor: Belen Diezma Iglesias UPM
  • Autor: Ramón Blanco Universidad de Zaragoza
  • Autor: F. Javier García-Ramos Universidad de Zaragoza
Grupos de investigación, Departamentos, Centros e Institutos de I+D+i relacionados
  • Creador: Grupo de Investigación: LPF-TAGRALIA: Técnicas Avanzadas en Agroalimentación
  • Departamento: Ingeniería Agroforestal