Mango Species Prediction from Leaf Using Convolutional Neural Network

Abstract

Convolutional Neural Network (CNN) is a vast area of researches in Machine Learning. Various types of researches are using Artificial Intelligence (AI) in agricultural fields, medical sectors for detecting diseases. This research is going to recognize mango species from the image of mango leaves. The most recent age of convolutional neural systems (CNNs) has gained exceptional outcomes in the field of picture grouping. This examination is connected with another way to deal with the improvement of mango species identification model, in view of leaf picture arrangement, by utilizing Deep convolutional neural systems (DCNN). We use DCNN for classification of mango species and detect them. Here we use images of mango leaves as our dataset. We have five classes and about 600 images as dataset. After classifying we train and test our system using Convolutional Neural Network (CNN). The accuracy of our system to detect mango species is 75%. It should be improved in further research.