Wearables can be used your young children or elderly persons to monitoring their locations or health, and one use case, especially for old age persons, is to detect falls. However, it’s quite possible they don’t like it and/or not always wear it, so the Center for Eldercare and Technology of the University of Missouri designed a system based on Microsoft Kinect, two webcams, and microphones in order to detect falls, and even predict falls by analyzing gait, i.e. the pattern of movement of the limbs.
Fall detection algorithms are relying on the microphone array, Microsoft Kinect depth camera, and a two-webcam system used to extract silhouettes from orthogonal views and construct a 3D voxel model for analysis. Passive gait analysis algorithms are for their part taking data from the kinect and the two-webcam system. The system was installed in 10 apartment, with data gathered for a period of 2 years, and they found that a gait speed decline of 5cm/s was associated with an 86.3% probability of falling within the following three weeks, and that shortened stride length was associated with a 50.6% probability of falling within the next three weeks.
You can see Gait detection in action in the video below.
More details about the studies and links to research papers can be found on Active Heterogeneous Sensing for Fall Detection and Fall Risk Assessment page on the University of Missouri website.
Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.