SpeciSphere Classification: An App that classifies an Image and Provides Veterinary Diagnostics for Canines (Dogs)

The paper discusses an intelligent image diagnosis and classification system that can identify general objects and make specific disease prediction in dogs. It starts by classifying the webcam-captured or uploaded image with a trained ResNet-50 convolutional neural network. Once it identifies the presence of a dog, it goes ahead to detect the breed and enables a rule-based diagnostic module to predict probable diseases, focusing on breed-related conditions. The diagnostic engine is based on an organized database correlating symptoms and breed predispositions to prevalent dog illnesses, facilitating more precise and context-dependent diagnosis. The web app, implemented with Python and Streamlit, supports real-time webcam functionality, user login, and safe activity logging through SQLite. The modular design of the system ensures user-friendliness, flexibility, and realistic applicability for pet owners, vets, and scientists. Possible enhancements in the future can involve integrating image-based disease detection and explainable AI methods to enable more automated and scalable diagnoses.