Human Detection and Tracking

Humans can enter the ViRoom and move around. First, the system will detect that someone appeared in the scene. After this detection, the human will be tracked in all cameras where he/she is visible.

Position in multiple images can be used to get a 3D position. And the other way round, this 3D position can be used in the case of uncertainty about the object position in a particular image.

Later, we will want to recognize human actions. Just knowing that a certain space is occupied by a human isn't sufficient for this task. Therefore, the system will find the head, hands and legs of the human and make his/her simplified stick model.