Alarm Filtering
Reduce noise and focus on real threats using AI-driven and threshold-based filtering.
Overview
Alarm Filtering prevents system fatigue by suppressing false triggers and blocking malfunctioning devices. This is achieved through AI object classification and "Overflow" protection.
Filtering Mechanisms
1. AI-Driven Filtering (False Alarm Filtering)
The system uses AI to analyze images/video associated with an alarm to distinguish between "Real" and "False" events.
- Priority List: Objects (e.g., "Person") that always trigger a real alarm.
- Whitelist: Potential objects (e.g., "Car") that may trigger an alarm based on location/time.
- Blacklist/Reject: Objects (e.g., "Bicycle", "Animal") that are suppressed as false alarms.
2. Threshold Filtering (Overflow Blocking)
To protect the system from malfunctioning devices (e.g., an Axon camera sending 100 motion events a minute):
- Default Threshold: 25 alarms per period.
- Blocking: The system automatically blocks a device if it exceeds the threshold, requiring an operator to "Unblock" it after the issue is resolved.
Configuration Steps
- Enable AI Filtering: In the GCXONE Configuration App, enable "False Alarm Filtering" for supported cameras.
- Review Analytics: Use the Analytics dashboard to verify if filters are correctly identifying objects.
- Adjust Thresholds: Modify the overflow custom property for sites with high legitimate activity.
Need Help?
If you're experiencing issues, check our Troubleshooting Guide or contact support.