Raven (Video Analytics)


Raven adds intelligence to the camera by analyzing video content to detect & determine temporal and spatial events in real-time. Raven is backed by state of the art deep learning models which captures & analyses the moment as is.

Abnormal Event Detection & Classification

Automatically monitors, detects and classifies abnormal events with a bounding box around that abnormal activity from camera feed and raises alerts in real-time and stores the clip of abnormal activity which can be viewed later. Events can be classified into:

  • Arson
  • Explosion
  • Fighting
  • Road Accident
  • Robbery
  • Shooting
  • Vandalism

Person of Interest (POI) tracking

Raven detects faces and generates high dimensional facial features from deep learning model and searches in POI database in real time.

Perimeter Breach Detection

Raven enables user defined perimeter monitoring and raise alarms in case of a breach in real-time and tracks/displays the breach affected area.

Application Areas:

  • Government Buildings
  • Airports
  • Railway Stations etc.
  • Defence, Police, Factories

Vehicle Counting & Tracking

Raven offers counting and tracking of vehicles at the intersection of roads in real-time. It can automatically detect the intersection of road or user can manually draw the intersection boundaries. It counts number of vehicles going straight, right, left, taking U-turn and the type of vehicle.


Does it take sound into account for video content analyses?

No, it does not take sound into account, currently it’s taking pure video for analysis.

Does the solution works in dark environment?

It works in dark as well light environment. The provided scene is visible.

Can it be deployed in on premises?

It can be deployed on edge, on premises & on cloud.