Abstract
Visual (VIS) and near-infrared (NIR) spectroscopy was applied to discriminate bruises and non-bruised healthy spots on ‘Golden Delicious’ apples. Two types of bruises were examined; those created by controlled impact and those by compression. Reflectance spectra of apples were measured in the range from 400 to 1700 nm. The data were analysed with canonical discriminant analysis (CDA). The squared canonical correlation (CR2) was 0.74 for discriminating impact bruises and non-bruised tissue, and a CR2 of 0.68 was obtained for distinguishing compression bruises and sound tissue. Based on the linear discriminant functions, built with canonical components, the misclassification errors for non-bruised apples were mainly due to the presence of compression bruises. The classification accuracy was improved by taking the type of bruises into account in the model.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
