Qualitative and quantitative evaluation of pomelo maturity using multivariate combination of chemical and physical properties
2009
TERDWONGWORAKUL, ANUPUN | PUANGSOMBAT, ARTHIT | PATHAVEERAT, SIWALAK
The present study reports the use of multivariate data analysis in combining physical and chemical properties for qualitative and quantitative evaluation of pomelo maturity. Pomelos of "Thong Dee" variety were harvested at four stages of maturity: immature, early mature, late mature and over-mature. The categories of maturity were classified with discriminant analyses on all variables and only nondestructive variables. The accuracy of classification using nondestructive variables (80.0%) was comparable to that using all variables (85.0%). For the estimation of the ratio of soluble solids content to acidity designated as a maturity measure, partial least squares regression models were developed based on either all remaining variables or nondestructive variables. The predictive performances by the two models were not significantly different. The acoustic response characterized by the third resonant frequency was shown to be the most important variable in qualitative and quantitative assessment of pomelo maturity. The results of the current study present a number of nondestructive parameters that are associated with pomelo maturity. They can be used to develop an on-line grading machine for pomelos.
显示更多 [+] 显示较少 [-]