Dragon Fruit Maturity Grading using Deep Learning Model

Grading of dragon fruit and the quality of product being delivered to the market is an important phase. The traditional methods are manual inspection, which is tiresome and damaging to productivity, because they are not producing consistent results. This paper postulates a deep learning-oriented approach for classifying the ripening stages of dragon fruit. We use a deep learning-based method in this scenario for classifying maturity stages of dragon fruit. It is then a matter of training the model using Convolutional Neural networks (CNN) on a dataset of dragon fruit pictures and classifying them as mature, not mature and spoiled. The proposed systematic method can be utilized for increasing the throughput of grading the fruits because it minimizes the errors and burden on the human.