Magnifera Indica cv. Harumanis Classification Using E-Nose

摘要

The Harumanis mango (Magnifera Indica cv. Harumanis) is a special variety with an exceptional commercial demand, and thrives well in the northern state of Perlis, Malaysia, In order to ensure that the fruits reaching the customers are of high quality, an investigation to automate the maturity and ripeness classifications of the Harumanis, using electronic nose with artificial neural networks (ANNs) has been conducted. In this work, a commercial e-nose, the Cyranose320 from Cyrano Sciences Inc., has been utilized. It is a robust E-nose with 32 conducting polymers carbon-based sensors which respond to most of the volatiles and hydrocarbon compounds emitted by the ripening mangoes. To further enhance the ANN performance, dimensional reduction and data manipulation were also applied. The results obtained are very encouraging and prove that classification based on the odour of the mangoes at different stages of ripeness is possible.