An Integrated Mask R-CNN and Domain Aware RAG-Enabled LLM Framework for Livestock Disease Detection and Care Recommendation

Livestock diseases significantly impact agricultural productivity, food security, and rural livelihoods, especially in regions like Nigeria where livestock farming is vital.Traditional diagnostic methods are often inaccessible or delayed in resource-limited settings, leading to unchecked disease spread and economic losses.This paper proposes an innovative framework combining deep learning techniques with a care recommendation system to facilitate early and accurate detection of prevalent livestock diseases such as Foot and Mouth Disease (FMD), Peste des Petits Ruminants (PPR), Bovine Fasciolosis, and Tick-borne Diseases.Using a comprehensive dataset of over 20,000 labeled images collected from veterinary sources and farms, a Mask R-CNN based model is designed to identify species-specific disease symptoms in cattle, goats, and sheep.The system integrates with a cloud-based large language model leveraging retrieval augmented generation (RAG) to provide tailored, actionable care advice, including treatment, and preventive measures.This approach aims to empower farmers with timely diagnostics and management strategies to mitigate disease impact, thereby enhancing livestock health, productivity, and overall food security.