Tender Coconut Classification using Decision Tree and Deep Learning Technique

A number of industry sectors such as Food & nutraceuticals, Pharma, Irrigation, Transport and Automobile use coconut and its parts for varied products. As different sectors make use of different type and parts of coconut, an automated separation of coconut Would aid in faster processing. Thus the major issue faced by these sectors is the automated sorting of coconut into different categories such as Non-Tender Coconut, Tender Whole coconut, Tender coconut cut With flesh and Without flesh. In this Work We have trained and tested a Decision Tree and DenseNet model to automatically classify coconut images into various categories. For Decision Tree 93% of accuracy Was obtained While the advanced Deep learning model, DenseNet, gave 100% accuracy. This showed that the Deep Learning models such as DenseNet perform better as compared to conventional machine learning algorithms such as Decision Tree.