Ensemble Learning of Maize Leaves Infected by Fall Armyworms Using CNN

One of the most commercially successful cereal crops farmed worldwide is maize (Zea mays L.), which is a common staple food in many developing nations. Massive yield loss is caused by the extreme outbreak of fall armyworm in maize. The purpose of this research is to develop a deep learning-based model that is trained on images from healthy and armyworm-infested maize leaves and to use convolutional neural network (CNN) architecture to increase accuracy using ensemble learning. Using the Keras functional API, the Ensemble model is trained on sequential model and two functional models that are ResNet-50 and VGG-16 to achieve a higher validation accuracy of 98.58%. Therefore, the detection of fall armyworm-infested maize leaves and the treatment of fall armyworms may result in an increase in crop production.