Aggregated Framework for Surveillance Camera Collaboration

Monitoring Crime through surveillance cameras has proved to be effective, however, it needs to be constantly improved. It is crucial to enhance the techniques used in crime monitoring through camera surveillance. In the current surveillance cameras, there are issues of redundancy and blind spots due to overlapping of cameras, lack of camera collaboration, and lower detection accuracy. To address these issues, a framework for a unified or aggregated approach is proposed, whereby cameras use vision controllers to properly track moving objects. Each camera is equipped with a vision controller, enabling it to tilt and turn in the direction of the target that is being tracked. In this paper, a scenario with 4 cameras and one person is used to show how the cameras work collaboratively. When the target moves away from the view of one camera, it will be covered by the next camera on the network. At the end of the day, the footage will be put together to have one footage that covers the entire range. Reinforcement learning is used so that cameras can learn from their own experience. The experimental section of the paper demonstrates that the proposed system outperforms other systems at its level.