Classification of Infectious Diseases in Chickens Based on Feces Images Using Deep Learning
Moch. Kholil, Heri Priya Waspada, Rafika Akhsani
Abstract
Artificial intelligence technology in deep learning is one of the popular classification methods. The development of deep learning technology is expected to assist farmers in identifying the types of infectious diseases that attack chickens based on feces images so as to increase production yields. Several infectious diseases that attack chickens can be identified through their feces, including newcastle disease caused by a virus, pullorum caused by bacteria, and coccidiosis caused by parasites. To identify, it is necessary to classify the types of diseases that attack by using images of chicken feces. With deep learning based on Convolutional Neural Network (CNN) in Keras/TensorFlow, the percentage of images that were accurately predicted according to the classification of infectious diseases suffered by chickens was 95.28% and less accurate was 4.72%.