Selected models of the autonomous 6000 series NOVUS recorders marked additionally with the letter F have the function of face recognition, which is related to the use of an appropriate signal processor. This function is available for the above-mentioned recorders and cameras from the 3000 series with the value of firmware 4.3.0 also marked with the letter F.
Faces are searched in the video stream and then retained in the database of recognized faces on the recorder. In addition, if the camera has a slot for SD cards, it is possible to record images of recognized faces and a full image of the scene in which the face was detected. In the case of connection to the camera through the IE browser, it is possible to write to the local hard disk in parallel. Face recognition can trigger an alarm action - trigger the alarm output, take a still image, save a moving image, send an email with a photo or make a recording to an FTP server. In order to obtain greater reliability of the performed diagnoses, the function can operate according to a specified schedule and in a limited field of view.
Very often, the system installers use the corridor mode, available in NOVUS cameras, which also has a face recognition function. This allows for more efficient use of the transmitter and is especially recommended for narrow passages through narrow doors.
In standalone recorders, the face recognition function is available for all channels of the recorder, so it can operate in parallel with up to 32 cameras (!). Its operation is configured from the recorder level and it is not necessary to connect to each of the cameras via the IE browser. The default camera operation mode is continuous operation.
Face recognition in NOVUS systems is constantly evolving. Currently, the functionality has been extended by adding an external database of identified faces and tracking the recognized face on the site map.