Anomalies Detection For Surveillance is a automonus system that aims to detect potential anomalies in an environment and make relevant decision to address that anomaly. We detect four potential anomalies weaponry, smoking, violence and fire. In case of these potential anomalies we make detection and perform relevant actions.
Background
In the market, currently systems are present which can detect the activities but those are very specific to a single activity. If we specifically talk about the surveillance system, consist of only the biometrics and cameras with DVR.
Those system does not involve any type of computer vision on real time.
But this software can detect the multiple type of anomalies that may be prohibited in the premises
of campus like Violence, Weaponry, Fire and Smoking .This system investigates the frames coming from cameras and look for the anomalous activities.
How the System Works
This system is the combination for multiple objectives:
i. Receiving the data from system
ii. Investigate the frames coming from specified cameras for anomalies.
iii. Detection and Localization of Potential Anomalies
iv. If necessary, system can raise the alarm in case of emergency.
Significance of the System
Specific implementation of this system is at the Educational Institute campuses where the above-mentioned activities are considered as anomalies.
But we can implement this in the areas like offices, building or factories. With very few changes in the model, we can implement this in the other areas as well for surveillance.
We can even use this in this pandemic situation where mask is mandatory to wear. We can detect and identify the persons with no mask.
