AI-Based Early Detection System for Dry Eye Disease with Automated Humidifier Control

Dry Eye Disease (DED) detection in current clinical practice relies heavily on manual diagnosis, patient-reported symp- toms, and periodic eye examinations, often requiring specialized equipment and trained personnel. Existing wearable solutions, while effective, are expensive and impractical for continuous, everyday use. This work proposes a cost- effective, non-invasive, and scalable system for real-time DED monitoring using a standard webcam and computer vision algorithms such as MediaPipe Face Mesh and Eye Aspect Ratio detection. The system continuously analyzes blink patterns to identify abnormal blink rates indicative of dry eye conditions. Upon detection, it automatically ac- tivates a connected humidifier to restore optimal air moisture levels, providing preventive intervention. Designed for seamless background operation, the solution is suitable for home, office, and multi-user environments, offering an affordable alternative to existing methods while integrating health monitoring with environmental control.