Artificially Ripened Mango Fruit Prediction System Using Convolutional Neural Network

India is one of the chief producers and consumer of mangoes. Mango being a climatic fruit, there is a high demand during the season because of which mangoes are usually ripened artificially with artificial ripening agents like calcium carbide. The artificial ripening agents cause lots of health hazards on the consumer because of which it is important to know if the mango fruit is artificially ripened or naturally ripened. Whereas, computer vision-based techniques are evolving for quality assurance, classification, defect, and disease detection of fruits. Hence, detection of artificially ripened mango fruit helps consumer in quick decision-making when compared with manual identification. The most successful deep learning model is convolutional neural network (CNN) which has made a remarkable achievement in the field of identification, defect detection, and classification of fruits. This paper proposes the usage of CNN-based artificially ripened mango fruit prediction with binary cross entropy for loss reduction. This model results in classifying artificially ripened mango fruits with increased rate of accuracy and better loss reduction. It has a good outlook with the artificially ripened mango fruit prediction system.