A feasibility study of using an Electronic Nose as a fruit ripeness measuring instrument

Abstract

This paper presents the study of using an artificial olfactory system as a nondestructive instrument to measure fruit ripeness. The cultivar chosen for this study is Harumanis mango. This system comprises of an array of semiconductor gas sensors as well as data acquisition and analysis components. It uses readings taken from Harumanis mangoes of different ripeness over a period of time, because at each stage of ripeness, mangoes will leave a different pattern or fingerprint onto the sensors array Readings taken from Harumanis mangoes of different ripeness over a period of time are used to train the system. Each stage of ripeness of the mangoes leaves a different pattern or fingerprint onto the sensors array. Artificial Neural Network (ANN) is then trained to classify the data into several stages of mango ripeness. The trained network is integrated into the system to allow mango ripeness recognition. Artificial Neural Network (ANN) is used to classify the data into the different stages of mango ripeness. The trained network is integrated into the system to allow mango ripeness recognition.