Determination of Chopped Fruits Freshness with High Accuracy by Using Electronic Nose

Abstract

Especially for the last 20 years many studies have been made in the field of health, chemistry and food with the electronic nose which is a very popular research area thanks to development of sensor technology. In these studies, high achievements have been obtained in many subjects ranging from disease detection to identification of bacterial species and from determination of food quality to aroma separation. In the study, the variation of the freshness of some fruits prepared for eating was examined with an electronic nose according to the days. Fruits as chopped melon, peach, banana and uncut strawberries were put on plates separately and all of them were stored at the room temperature for 5 days. A low cost electronic nose has been made with the MQ branded 11 gas sensors. Fruits were sniffed 15 times for each day at the same time zone. Sensors' data were transferred to the computer via 2 Arduino Uno microcontroller cards and recorded to computer with an interface program made up in the LabVIEW environment. Then, features were extracted from the obtained data taken from the sensors, and the extracted features were classified by using the k-Nearest Neighbors and Neural Network Classification Algorithms. Best classification accuracies were obtained as; 93.28% for melon, 80.80% for peach, 84.80% for banana and 75.78% for strawberry.