It is likely that if you get a security camera system, that you will want to be able to identify an intruder that has been captured on security camera footage. This process is called facial capture. Some people also call it facial recognition although they are really two different processes. Facial capture means to get enough usable details to be able to identify the person of interest by comparing facial features. It doesn’t mean the security cameras will identify the person, only that enough detail has been captured that he can be identified. Facial recognition actually refers to analytic software that can compare and match captured images of a persons face against another in a database. There are many interesting applications for such a process, but that is an entirely separate article.
Getting proper facial capture is an important process to understand because what good are your security cameras if you cannot identify the person of interest in the video? Proper identification requires at least 25% of the video image to be occupied by the face. This means that you really need to use a dedicated camera for this process. Begin the installation by finding an ideal spot to capture the face. Usually an entrance or exit would be ideal, but maybe there is a room entrance to something valuable. Once you have found the perfect spot to capture the face details, decide on the type of camera you will use for this process. It is best to use a good quality day/night security camera with a vari-focal lens rather than an infrared camera. Infrared cameras will be taking video in black and white in low light and this could make identifying the intruder more difficult. Since you will be using a good low light camera, make sure there is at least a little bit of ambient lighting. A night light, motion light or something of the like will be fine. Mount the camera close enough to the capture point so that at least 25% of the video (preferably much more) is the subjects face. Make sure to accommodate for people of different heights. Next, mount a second camera in a corner, or somewhere else in the room to provide you with a good overview. Now you have one cameras that is used to only capture facial details and a second camera that is providing the general overview of what is going on in the room. Make sure that the channel that the facial capture camera is on is set to record at the highest resolution possible. A D1 resolution image will be 4 times larger than a CIF resolution image and therefore will provide you with 4 times more detail. Use the higher resolution setting even if the frame rate has to be reduced. The overview camera can be set at CIF resolution rather than D1.