Predicción de los sólidos solubles totales, ph y acidez titulable de naranjas (citrus sinensis l. var. valencia) mediante imágenes hiperespectrales

• Autor: Aredo Tisnado, Víctor Jesús
• Formato: tesis de grado
• Fecha de Publicación: 2015
• Institución: Universidad Nacional de Trujillo
• Repositorio: UNITRU-Tesis
• Lenguaje: español
• OAI Identifier: oai:dspace.unitru.edu.pe:20.500.14414/4435
• Enlace del recurso: https://hdl.handle.net/20.500.14414/4435
• Nivel de acceso: acceso abierto
• Materia: Predicción, Mínimos cuadrados parciales, Imágenes hiperespectrales, Atributos de calidad, Naranjas.

Hyperspectral imaging in the visible and near-infrared (400–1000 nm) regions was tested for nondestructive determination of total soluble solids (TSS), pH, and titaratable acidity (TA) in oranges (whole and half) in commercial ripeness. The spectral data were analyzed using the partial least squares (PLS) analysis. The determination coefficients (R2) with the whole spectral range (400–1000 nm) for predicting TSS, pH and TA on whole oranges were 77.0%, 77.2% and 78.3% with Standard Error of Calibration (SEC) of 0.501 ºBrix, 0.080 and 0.092 % citric acid, and Standard Error of Prediction (SEP) of 0.517 ºBrix, 0.080 and 0.088 % citric acid, respectively; for half oranges the R2 for predicting TSS, pH and TA on half oranges were 92.1%, 87.7%, y 88.0% with SEC of 0.294 ºBrix, 0.059 and 0.068 % citric acid, and SEP of 0.400 ºBrix, 0.061 and 0.074 % citric acid, respectively. The most influential wavelengths were selected using coefficients β from PLS models. New simplified PLSR and multiple lineal regression (MLR) models were established using only the selected wavelengths to predict the quality attributes, but the models didn’t have an acceptable level of prediction (R2<70%).