영문내용
(English Abstract) |
Fall detection is a major challenge in health care domain, especially important for the elderly; falls can make serious injuries to them, worst case may lead to death if no early aid is provided. Therefore, an automatically fall detection system is critical. We will propose work-in-progress of a real-time method for fall detection system based on only movements. Our system will be trained with normal activity and have the ability to detect abnormal movement in run-time, then combine with the length of stationary to detect falls. If a vast and abnormal movement is detected followed by a long period of lack of movement, then an alarm will be triggered.
|
