This project presents an AI-enhanced skin screening system implemented on an edge device, specifically designed for portable, real-time detection of dermatological conditions and concurrent physiological monitoring. Our system leverages a compact processing unit integrated with an imaging module for continuous facial image acquisition. These images are processed by a pre-trained Convolutional Neural Network (CNN), enabling the classification of skin health and the identification of conditions such as keratosis, basal cell carcinoma, melanoma, benign lesions, and nevus. Extending beyond visual screening, the system incorporates physiological sensors for continuous measurement of vital parameters, including heartbeat and temperature. Detection of anomalous fluctuations in these vital signs triggers an immediate auditory alert and a visual display of the detected skin status and physiological readings. This integrated solution demonstrates the feasibility of real-time, non-invasive skin anomaly detection and comprehensive health monitoring on a compact, low-power edge device, offering convenient, accessible, and early-warning capabilities for personalized health management.
PRASAD et al. (Wed,) studied this question.