An Automated Alert System for Monitoring the Hygiene in Restaurants Using Machine Vision

Hygiene violations in restaurant kitchens pose significant risks to public health, necessitating effective monitoring and compliance systems. This paper introduces a sophisticated system integrating machine learning, computer vision, and cloud technologies to monitor and enforce hygiene standards in restaurant kitchens. The system uses a YOLOv8 machine learning model trained on a specialized dataset to detect hygiene violations such as improper attire (apron, gloves, hairnet), presence of pests (lizard, rat, cockroach), and adherence to hygiene protocols. Real-time monitoring is facilitated through connected cameras strategically placed in kitchen areas, allowing instant detection and notification of violations. The system's capabilities are improved by the integration with Firebase services, which include Cloud Storage and Database. The Firestore database securely stores violation details and sends out real-time alerts to ensure prompt resolution of such problems. The system achieves an impressive accuracy rate of 89% in identifying and categorizing hygiene violations, ensuring adherence to food safety regulations and safeguarding public health in restaurant environments.