Abstract | ||
---|---|---|
A method for sending real-time fall alerts containing an embedded hyperlink to a depth video clip of the suspected fall was evaluated in senior housing. A previously reported fall detection method using the Microsoft Kinect was used to detect naturally occurring falls in the main living area of each apartment. In this paper, evaluation results are included for 12 apartments over a 101 day period in which 34 naturally occurring falls were detected. Based on computed fall confidences, real-time alerts were sent via email to facility staff. The alerts contained an embedded hyperlink to a short depth video clip of the suspected fall. Recipients were able to click on the hyperlink to view the clip on any device supporting play back of MPEG-4 video, such as smart phones, to immediately determine if the alert was for an actual fall or a false alarm. Benefits and limitations of the technology are discussed. |
Year | DOI | Venue |
---|---|---|
2014 | 10.1007/978-3-319-14424-5_5 | Lecture Notes in Computer Science |
Keywords | Field | DocType |
Fall detection,Fall alerts,Kinect | World Wide Web,False alarm,Computer science,Hyperlink,Multimedia | Conference |
Volume | ISSN | Citations |
8456 | 0302-9743 | 3 |
PageRank | References | Authors |
0.45 | 8 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Erik E. Stone | 1 | 381 | 31.42 |
Marjorie Skubic | 2 | 1045 | 105.36 |