Activity Recognition for Healthcare Based on Slow Intelligence Systems
Wenhui Chen, Bing-Yang Chang · 2016
Activity recognition has many potential applications, ranging from healthcare, security monitoring to human-computer interaction. The deployment of sensors is required either in the human body or in an environment where the subject lives to identify activities. However, due to the fact that sensors are subjected to a variety of errors and the nature of human activities are dynamic and uncertainties, it is a challenging task to develop an accurate activity recognition system merely with sensor readings. In this study, we present a novel approach to activity recognition with scene information captured from a wearable camera. The scene information is obtained by analyzing image inputs using speed-up robust features. With the scene information, some unlikely activities can be ruled out to help identify the performed activity. Our experiments showed that the recognition accuracy can be increased from various test scenarios.