Fruit Recognition Using Deep Learning Approach

Abstract

The paper intends to develop a model for the identification and classification of fruits using the concept of deep learning. The goal is to create an automated feature extraction system using convolutional neural networks. Fruit recognition is used in agricultural applications such as robot harvesting and fruit counting, and many more applications. The system is capable of sorting the fruits. It can be used to inspect the state of fruits and determine whether they are fresh or not. Fruits can be identified using the self-service system in the retail store. The suggested method takes advantage of the 'ImageNet' dataset, which is of high quality. The fruit images in the collection are divided into five groups. Dataset is very challenging as the images contain diverse fruits of the same colour, shape and fruits are overlapped. Convolutional Neural Networks are used in the model to recognize fruits from images. The accuracy was found to be 92.23%. Machine learning methods are outperformed by deep learning algorithms.