The Development of Chicken Disease Detection Application Using the Xception Transfer Learning Method
Vitri Tundjungsari, Shela Atya Mitasya · 2024
Chicken farming in Indonesia is a potential business, considering the high demand for chicken meat and eggs. However, this business also has a significant risk of loss due to diseases that often attack chickens. This problem leads this research to find a solution so farmers can reduce the risk of loss and handle chicken diseases more effectively. To answer this challenge, we develop an application specifically designed to help chicken farmers recognize various types of chicken diseases and the symptoms that appear and recommend preventive and treatment actions. This application also has a direct disease detection feature that allows farmers to get recommendations for the right medicine and the nearest purchase location. To develop the application, we use the Xception Transfer Learning Method. Xception is a deep learning model developed by Google to improve the Inception architecture, tailored explicitly for computer vision tasks like image classification and object detection by reducing the number of parameters, making the model more efficient without losing accuracy. The testing result shows that by utilizing transfer learning, the model successfully accurately classified various types of chicken diseases based on feces images. This application is expected to be an effective tool for farmers to minimize the risk of loss and maintain the health of their livestock, especially in Indonesia.