Poultry Disease Detection using Deep Learning -An Ensemble Model Approach

This paper appears a computerized poultry disease detection framework utilizing profound learning and picture handling methods. The framework centers on recognizing three basic poultry diseases—Coccidiosis, Salmonella, and Newcastle Disease—through investigation of fecal pictures. The mobile-based usage gives an open arrangement for ranchers without requiring specialized gear or ability, giving this important for creating districts where veterinary administrations may be limited. This paper shows a computerized poultry contamination area system utilizing significant learning and picture taking care of strategies. The system centers on recognizing three fundamental poultry diseases—Coccidiosis, Salmonella, and Newcastle Disease—through examination of fecal pictures. Utilizing a dataset of 8,067 poultry fecal pictures over four categories (counting solid tests), we executed and gathering making together MobileNetV2 arrange with custom CNN models. The framework fulfills 92.07% accuracy in disease classification, essentially outflanking standard visual overview strategies.