User:StanGanin

 Elderly falling detection 

Analyzing the body position and movements of elderly people is an important application of computer vision. By using video surveillance it is possible to automatically monitor the movements of elderly people and detect if abnormal behavior occurs, i.e. if the person falls. Since falls are a big risk in the elderly, such a system could substantially improve the quality of their independent life by providing a safer environment - by reducing the response time of medical personnel and reducing medical costs.

Existing solutions
There are a few options for the types of devices used for fall detection. Wearable device-based - having devices with gyroscopes and accelerometers on the person can give real time information his posture, position and acceleration. The other type is ambience device-based. The third method is based on computer vision. This has quite a lot of advantages over the other two options, the main of which is that it does not interfere with the personal privacy of the person. Another big plus is that the data gathered with this method can be collected and analyzed for further improvements.

Using a Time-of-Flight 3D camera
Time of Flight cameras is a range based camera. It calculates it based on the speed of light which is known by taking into account the time that is needed for the light to travel to the subject. These cameras have advantages for this kind of usage because they have small dimensions and lower power consumption. Also they are accurate enough for detecting falls and motion, but not accurate enough to detect facial features, thus keeping the privacy of the monitored person safe.