Smart Video Analytics
Our current world is moving faster and with more information than ever. Managing and understanding this has become a challenge that can only be mastered with the help of smart systems utilizing machine learning (ML) and artificial intelligence (AI) on-board.
The increasing ubiquity of video surveillance systems and the inherent amount of information these systems carry prohibits train operating companies to proactively extract all the available insights and information from these systems.
ML and AI algorithms
Increasingly performant ML and AI algorithms allow train operating companies to capture this information in real-time. Moreover, the rise of edge-based ML and AI algorithms remove the need for high bandwidth and highly reliable connection with the cloud.
Video surveillance system are recording more video footage than anyone could ever view and analyze, hence utilizing ML and AI algorithms to detect various conditions, is gaining traction.
Detecting camera tampering, left object detection, seat occupancy or passenger crowding in certain areas, or detect people running down an aisle, or even fire detection have become a reality.
Many more functions are being developed and improved through collaborations with our customers.
- Detect events in live video footage.
- Create a special event video file beginning before the event.
- Have easier access to forensic footage.
- Alert staff of event, even on the wayside.
- Camera tampering detection.
- Running detection.
- Left object detection.
- Crowding estimation.
- Fire detection.