Integrating MAX30100 Sensor and MediaPipe Hand Gesture Recognition for Real-Time Health Monitoring System.
Shrinidhi Venkatesh Bhat, Mansi Malayka, Aparna Mohanty · 2024
This paper presents a sophisticated framework for real-time health monitoring and gesture-based interaction, integrating advanced technologies like the MAX30100 sensor, Arduino microcontroller, convolutional neural networks (CNNs), and a secure database infrastructure. The system enables continuous monitoring of vital signs such as heart rate and oxygen saturation levels (SpO2), while also allowing intuitive user interaction through hand gestures captured via live video feed. Central to this framework is a carefully designed hardware setup combining the MAX30100 sensor and Arduino microcontroller, ensuring precise measurement of physiological parameters. Simultaneously, a CNN-based gesture recognition model, supported by the Temporal Shift Module (TSM) and 2D CNNs, accurately identifies and categorizes various hand gestures in real-time for seamless user-command interpretation. Data acquisition and management are meticulously handled, with raw physiological data and gesture recognition outputs seamlessly integrated into a secure database infrastructure. This repository serves as a real-time reservoir for health metrics and archives user interactions for further analysis and system optimization. Moreover, The framework integrates predictive modeling algorithms to anticipate health issues from historical data. It promptly alerts and transmits data to healthcare professionals upon anomaly detection or predicting impending health concerns. Beyond its technical capabilities, this framework embodies a holistic approach to healthcare technology, aiming to improve patient outcomes and quality of life through proactive monitoring and personalized care delivery. By combining advanced sensor technologies, machine learning algorithms, and user-friendly interfaces, it represents the convergence of innovation and effectiveness in modern healthcare systems.