Aglaonema or popularly known in Indonesia as "Sri Rejeki" is a leaf ornamental plant fancied by many people. This plant has unique leaves with beautiful and diverse shapes, colors, and patterns. Various ways can be used to identify this plant; one of which is by using an image processing technique in which the process is carried out through feature extraction or classification process. A method/algorithm to classify Aglaonema image is the Convolutional Neural Network (CNN). CNN is an algorithm of Deep Learning and is the development of a Multi Layer Perceptron (MLP). This study used the image of 5 types of Aglaonema leaves with 100 images of each type. The CNN model used in this study was the Alexnet model. Based on 4 experiments using the optimizer and different configurations of epoch values, the highest training validation accuracy value was 98.00%. The system also can classify Aglaonema images well with an accuracy success rate of 96% of 50 images tested.
Copyrights © 2022