Using the data on a person´s pose in the “smart home” system

The problem of improving the effectiveness of the through the automatic recognition of images of the human body is solved in this paper. There is the scheme of an event-efficient model that describes the general principle of the system “smart house”. A model apartment plan is studied, a table of events and activities for the premises is based on it. This table takes into account the position of a person, depending on which system performs certain actions. Also the existing system of recognition of human posture is described in the paper, there is a table, their comparative analysis, which made the choice in favor of system Kinect. The existing algorithms for determining the position of a person are studied. This system will provide new data about the status of the person, which will make possible to take a more flexible solution.