Identification of artificially ripen mango using aroma and texture features

Mango is known as a king of fruit. It has a unique sweet taste, flavor, and nutritionally rich attributes. Therefore, it is beloved by all age groups of people. Looking at the high popularity of ripening mangoes, vendors are attracted to artificial ripening using hazardous chemicals like calcium carbide. Most governments have banned the use of calcium carbide. Unfortunately, the use of calcium carbide still exists as per several reports. Today, the public is being aware and cautious to avoid buying such artificially ripened mango. But, it is difficult for humans to differentiate an artificially ripe and naturally ripe mango and that too specifically for some species like Kesar by the human eye. This paper proposes a non-destructive approach to identify an artificially ripened Kesar mango. The first phase uses image features and a decision tree to classify mango into ripe, unripe, and partially ripe categories with 93% accuracy. The second phase discards unripe mango images and processes to identify artificially ripened or ripening mango using GLSM based image processing and an MQ3 aroma sensor.