| Autor: |
Armeniakos, Charalampos K., Nikolaidis, Viktor, Tsekenis, Vasileios, Maliatsos, Konstantinos, Bithas, Petros S., Kanatas, Athanasios G. |
| Zdroj: |
Journal of Ambient Intelligence & Humanized Computing; Sep2023, Vol. 14 Issue 9, p11657-11669, 13p |
| Abstrakt: |
Accurate and timely human fall detection is a strong requirement either for the surveillance of critical infrastructures or for ships. Indeed, sea-faring vessels are one of the most important means for maintaining the marine economy in many countries by transporting goods or people. However, unfortunate tragic accidents on-board ships involving people, either a member of the ship's crew or a passenger who has fallen off the ship may take place, which is known by the term "man overboard" (MOB). Accordingly, the use of radar sensors for human safety monitoring applications is vital and is of special interest since it is proven that radar sensors are less influenced by environmental conditions (e.g. fog, rain, temperature) compared to other systems like video cameras. Consequently, human fall detection from either sea or ground infrastructures is easier to be identified using radars compared to the conventional methods. This paper focuses in the description of a real experimental approach based on multiple long-range millimeter-wave band radar sensors for human fall detection. The stream(s) of information collected by the system, are processed using clustering techniques. The clustering results are evaluated in terms of the ability to detect and track real human fall scenarios. The results reveal that the measure of velocity plays a key role in the detection procedure. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
Full text from SpringerLink