Fruit Ripeness Detection Using CNN Model

Abstract—Fruit ripeness detection is crucial in the agricul- ture field to determine the time of harvest and fruit quality. Substantial crop loss might result from timing slips or delays, consequently affecting income rates. Manual detection of ripeness in fruits can be inefficient for large scale implementation and pose issues such as requiring lots of time and intensive labor. Hence, in this project, fruit ripeness is detected utilizing computer vision and machine learning technologies. The method being used will recognize the fruits and categorize their ripeness using the CNN algorithm, on the basis of the dataset it is trained on. Our system will detect ripeness of 6 different types of fruits. The convolutional layers applied different filters as feature detectors to create a feature map representing the detected input features. Then the features were shaped into one-dimensional features and sent to an Neural Network for the ripeness classification. Data augmentation techniques are used to expand the dataset’s size prior to building a trained model. This technology can be integrated with robotics to implement smart agriculture and help farmers view their crops in real time, know the number of mature fruits, plan harvesting and assist in fruit picking and sorting. Index Terms—Automated Detection, Machine Learning, Deep Learning, Computer Vision, Agriculture 4.0, Smart Farming.