Wearable Sensors in Human Activity Recognition: A Survey

Abstract

The field of human activity recognition (HAR) has been revolutionized by wearable sensors; it has created a new frontier of opportunity in the fields of health monitoring and fitness training. This survey paper presents an overview of wearable sensors, focusing on design and placement, and issues in HAR. A historical development of wearable sensors in the context of HAR, focusing on key technological advances really shaping the field, is sketched out. The different sensors used are presented in detail: accelerometers, gyroscopes, and heart rate monitors, with comparative efficiency in recognizing various human activities. It also covers very important aspects of sensor design and placement, which affect highly the accuracy and reliability of HAR systems. The major data quality challenges, sensor calibration, and user compliance are described and discussed separately. Additionally, this paper proposes a system integrating multiple wearable sensors with an AI-driven data processing framework to enhance the accuracy and efficiency of HAR systems. Then the paper gives several applications of wearable sensors in HAR, from health care to sports and fitness. It concludes with new emerging trends: integration of artificial intelligence with wearable sensors and development of new sensor technologies, showing the way to exploit and merge new research and innovation.

Keywords: field, recognition, revolutionized, sensors, intelligence