Linear discriminant analysis based Indian fruit juice classification using NIR spectrometry data

摘要

A large variety of fruit juices are available in the Indian market. These juices differ from one another in terms of quality and added ingredients. Fruit juices from two different commercial brands are compared with their consequent homemade ones using a portable near infrared (NIR) spectrometer in the range of 250–925 nm of wavelength. Juices extracted from apple, grapes, mango, orange, pineapple and pomegranate are used as samples for our experiment. The results are clustered using linear discriminant analysis (LDA) for the visualization of class distinction. Despite variations in the contents of a particular fruit juice obtained from different sources, a particular pattern is observed. The characteristic features of the commercial juices and homemade ones for each fruit juice are visible in the near infrared range. Linear discriminant analysis (LDA) has rendered a distinct clustering pattern to successfully differentiate all commercial brands and homemade juice for each fruit.