Classification of Fruits using Convolutional Neural Networks

Abstract

Fruit classification and disease detection plays an important role in the intelligent agricultural farms. Fruit classification is critical in a wide range of industrial organizations, including factories, supermarkets, and other environments. The significance of fruit classification can also be observed among those with special dietary needs, who use it to assist them choose the appropriate types of fruits. Convolution Neural Networks (CNN) is one of the most advanced Deep Learning techniques, with image recognition taking the lead. We have supplied a dataset with a variety of fruits, and evaluated them based on pattern recognition. To produce the most refined prediction for fruit classification and disease detection, we used required convolution and pooling layers. When thoroughly analyzed by feature extraction and image segmentation, CNN demonstrated good accuracy as compared to other models. Our work is primarily focused on obtaining an classification of various fruits, the CNN model gives accuracy 98.6%.