AUTOMATED SYSTEM FOR CONTROL OF MULTIPLE CAMERAS WITH RESPECT TO THE ACTIVE SPEAKER

In this paper, the research on automated system for control of multiple cameras is presented. The principal goal is to enable switching of cameras based on the position of active speaker. For the purpose of this system it was crucial to design PTZ cameras positioned in specific location, i.e. local government hall. In total, three independent system components are created and communicate with one another using REST calls. Each of these components deals with a specific logic part of the whole system. The first component recognizes the active speaker by monitoring the input along using OCR system and consequently outputs name of the speaker. The second component is for the management of members stored in the system and accepting recordings of active users. The last component is the most user oriented. It has a graphic interface in which the user can define cameras position beforehand and manage members of the system. The most critical function is the automatic behavior of cameras.