Visual-based public safety event detection

• Only video analytics with embedded, on-the-edge hardware
• Public safety events create situational awareness
• Low-latency and real-time public safety events, accessible from any (VMS) platform
• Video is completely anonymized through privacy-by-design
• Detections robust against lighting conditions, appearances and installation angles

Input(s):

• Video stream from surveillance camera

Main feature(s):

• Detect events of loitering, crowdedness and running crowds to classify potential public safety hazards
• Monitor presence and counts of pedestrians and vehicles in embedded hardware
• Alarms can be used to inform police or other authorities through a user platform

Output(s):

• Public safety events over API
• Presence/counts over API

• Camera calibration is required for accurately calculate speeds
• Camera quality should be sufficient for local light conditions

Public safety observation rooms, crowd management offices, smart City Platforms (e.g. MyCity or Esri platform).

Commercial product.