Multi-modal Fall Detection within the WeCare Framework

Falls are identified as a major health risk for the elderly and a major obstacle to independent living. Considering the remarkable increase in the elderly population of developed countries, methods for fall detection have been a recent ac- tive area of research. However, existing methods often use only wearable sensors, such as acceloremeters, or cameras to detect falls. In this study, in contrast to the state of the art solutions, we focus on the use of multi-modal wireless sensor networks within the WeCare framework. Our set-up not only includes scalar sensors to detect falls and motion but also consists of embedded cameras and RFID tags and uses sensor fusion techniques to improve the success of fall detection and minimize the false positives.

Experiment Setup

In this study, we present a fall detection method using both wearable acceloremeter sensors and embedded cameras. We use the IMOTE2 wireless mote with ITS400 sensor board for acceleration. 2 IMOTE2 motes are attached on the subject's body, one on the trunk and one on the waist as shown in Figure 1 and AXIS 207W network camera is used for capturing video . In the demonstration, the user is asked to perform activities such as walking, jumping, lying down, sitting and quickly sitting which should be distinguished from a fall by the system. The performance of the system is evaluated by successfully detecting falls and by minimizing the false alarms (i.e., other activities reported as fall). An example snapshot from a set of evaluations is shown in Figure 1.

Figure 1


Accelerometer data is thresholded to detect major drops (Figure 2) and classify the activities. Video images are preprocessed to separate the blob corresponding to the human object from the background (Figure 3). An ellipse model is fitted to this blob. The eccentricity and orientation of this ellipse are used as features to classify falls. Accelerometer and video data are fused by using a decision fusion mechanism.

Figure 2

Figure 3

wecare-fall.txt · Last modified: 2010/09/20 09:43 by ozlem Creative Commons License Valid CSS Driven by DokuWiki do yourself a favour and use a real browser - get firefox!! Recent changes RSS feed Valid XHTML 1.0