Enhancing Poultry Disease Classification Using Fecal Image: A Fusion Approach

Abstract

The rising demand for animal products is causing the agricultural sector, specifically poultry farming, to increase its production capacity. This increase in poultry farming can raise the risk of spreading contagious diseases like Newcastle, Coccidiosis, and Salmonella which may result in massive death rates among chickens as well as other serious economic losses. For detecting these diseases traditional techniques are labor-intensive, time-consuming, and expensive. Additionally, there are insufficient expertly trained professionals in rural areas. A deep learning-based model is proposed to detect early classification of these diseases using fecal images. The proposed model utilizes a hybrid architecture combining EfficientNetV2B0 and DenseNet121 models for classification and achieves a high accuracy of 97.78% on the test set. Evaluation metrics including precision, recall, F1-score, confusion matrix, and ROC curve analysis illustrate the model's effectiveness in accurately categorizing poultry diseases. This approach offers a promising solution for early disease detection, enabling proactive health management in poultry farming to mitigate economic losses and safeguard human health.