Distributed Fog computing system for weapon detection and face recognition

Surveillance systems are very important to prevent situations where armed people appear. To minimize human supervision, there are algorithms based on artificial intelligence that perform a large part of the identification and detection tasks. These systems usually require large data processing servers. However, a high number of cameras causes congestion in the networks due to a large amount of data being sent. This work introduces a novel system for identifying individuals with weapons by leveraging Edge, Fog, and Cloud computing. The key advantages include minimizing the data transmitted to the Cloud and optimizing the computations performed within it. The main benefits of our proposal are the high and simple scalability, the immediacy of the detection, as well as the optimization of processes through distributed processing of high performance in the Fog layer. Moreover, the structure of this proposal is suitable for 5G camera networks, which require low latency and quick responses.