Camera selection and positioning
When selecting and positioning cameras, it is important to know the client’s surveillance needs. Do you need to detect objects/persons, to recognize people, or do you need to identify unique facial characteristics? The operational requirement determines the level of detail required.
To define recognition, this means that the image can be used to identify someone already known to the observer, even if the person’s face is not completely visible or clear - as other visual factors (height, body size, gait, clothes, etc.) also play a part in recognition.
An image good enough for identification however, means that there must be enough detail to allow positive facial identification of a previously unknown person, when this image is compared to other images or when the person is physically present, and regardless of any other considerations.
Although a single camera can provide an overall view of the scene, it might not provide enough detail to identify individuals. If this is one of the surveillance goals, then an additional camera needs to be included in the design, as in the figure below.
Figure 4 illustrates how four fixed cameras cover most of the parking area, whereas a single PTZ camera can cover the same area, by panning around. However, a PTZ camera can only view one segment of the area at a time, while the fixed cameras provide coverage all the time.