Prediction of sensory quality by near infrared reflectance analysis of frozen and freeze dried green peas (Pisum sativum)

Frozen, green peas (Pisum sativum L ) of different varieties and different levels of maturity were evaluated by sensory analysis using a panel of ten trained judges. Two texture variables (hardness and mealiness) and four flavour variables (pea flavour, sweetness, fruity flavour, off flavour) were considered. Near infiared reflectance ( N I R ) analysis was performed with the same material, on both the fiozen and fieeze dried peas. The N I R instrument was calibrated to predict the sensory variables using the multivariate analytical method of principal component regression. Tenderometer readings of the same peas were also calibrated to predict the sensory variables. N I R analysis on the fieeze dried peas showed relative ability of prediction (RAP) values for the sensory variables which were higher than those for the tenderometer readings. The sensory attributes pea flavour and hardness were predicted with higher RAP values by tenderometer readings than by N I R analysis on fiozen peas. For the rest of the attributes, N I R analysis on fiozen peas gave higher RAP values than tenderometer readings. N I R generally gave high RAP values, and this tentative study suggests that N I R analysis could be a useful tool in instrumentally assessing the quality of frozen peas.