Somebody call it Detector left and gone items.

If the standard video detector shows any changes in the frame compared to the previous one, the background detector gives static information, showing only changes relative to the background. 

They can be any fixed object that has changed its appearance or disappearance of any part of the background of the frame. For example, put on the floor bag, or carried away. Most often, such an algorithm is required by public authorities to detect unattended items containing explosives.

Unfortunately, the "left object" can be a body, and even a human leg, because it also stands still, obscuring the background, for example, while waiting for a train on the platform. And a new the glare of light, and opened the door, and many other things. Therefore, this algorithm did not have a successful application for a long time.  But everything changed with the advent of neural networks.

Speclab combines or in some cases replaces the background detector with a neural network that can recognize objects and classify them. The GOAL system automatically distinguishes between boxes, bags, backpacks and other items in which a bomb can potentially be stored.

Analytics SpecLab is also able to determine that the subject remained unattended: there is no other class of objects – people. Also, the GOAL system can show the person who left or took the item.