Ripening state determination of guava fruit (Psidium guajava) using e-nose with fuzzy logic as pattern recognition tool

Electronic nose is a non-destructive artificial olfactory system to sense aroma, can be used for classification of fruits at the time of harvest, post-harvest and during storage with various conditions. This paper is about development of application specific e-nose system with fuzzy logic as pattern recognition tool. The developed systemis used for classification and identification of guava fruits at the time of prepackaging.The developed system comprises of an array of eightSnO2MOS gas sensors, static odor delivery system, signal conditioning circuit,data acquisition and pre-processing software using LabVIEW2012. The design of this system focused on studying the response of a gas sensor array to various VOC vapors released by fruit during ripening and optimizing the data acquisition, signal preprocessing and pattern recognition. Fuzzy logic module is designed for pattern recognition. Classification results obtained from fuzzy logic module showed 90% general agreement with the results from thehuman classifiers. The results suggest that this e-nose system with fuzzy logic based pattern recognition is capable to classify guava fruits up to the expectation and it would be a feasible system to be used in a real scenario.