Abstract

This research proposes an automatic approach for classifying pigeon pox lesions using the DenseNet model, a deep learning architecture renowned for its dense connectivity patterns and efficient feature reuse. Pigeon pox, caused by the avipox virus, poses significant threats to avian populations and the poultry industry worldwide. Traditional diagnostic methods are laborious and time-consuming, prompting the exploration of machine learning techniques for automated disease diagnosis. The proposed approach leverages DenseNet for feature extraction from pigeon pox images, aiming to develop a robust and accurate classification system capable of distinguishing between healthy and infected birds based on visual features extracted from skin lesions. The study systematically evaluates the performance of the DenseNet model, demonstrating exceptional discriminatory power with an AUC of 0.99, accuracy of 0.98, and F1-score of 0.98. Comparative analysis with other deep learning models further validates DenseNet's superiority in pigeon pox classification. Overall, the proposed approach offers a promising solution for the early detection and control of pigeon pox, thereby improving bird health and welfare while mitigating economic losses in the poultry industry.