Patient Aid Message Notification System Based on Hand Movement Tracking and Haar-like Features

摘要

The human-computer interaction (HCI) has an important role in human lives for decades. One important area includes an assistive technology in medical system. An application such as patient monitoring is useful for both patients and caregivers. This work aims to help patient with speaking problem to communicate with other people without the need of using voice. This paper applies a combination of machine learning technique based on Haar-like features and a movement tracking method to identify direction of patient hand movement. Each moving direction is bound to a given message. The message is also displayed on a screen and sent to caregivers via a notification system. The experimental results show that the system provides satisfying outcomes in both recognition and movement tracking.