How mohitrawat5399-source/real_time_alert_system works
This is a Streamlit real-time person-detection alert system. A user configures monitoring and a YOLO model; the app captures webcam frames, runs detection, and treats person detections as alerts. It records alert details in session state, captures screenshot evidence, and updates the live view, logs, and gallery; audio is optional. The README describes configurable alert conditions generally, while the sampled code specifically shows person detection. Source excerpts omit some setup and alert logic, so uncertain wiring is not shown.
Main entry points: stream_lit.py Streamlit app · yolov8n.pt YOLO weights · config.json App configuration
Components 14
User interface
- Streamlit app
stream_lit.py - App configuration
config.json - Live dashboard
stream_lit.py
Detection
- Frame capture
stream_lit.py - YOLO weights
yolov8n.pt - YOLO detection
stream_lit.py - Person detection
stream_lit.py
Alerts and history
- Alert handling
stream_lit.py - Session alert history
stream_lit.py - Screenshot evidence
stream_lit.py - Alert log and gallery
stream_lit.py
Other components
- Operator
- Webcam
- System beep
Connections 14
- Operator → Streamlit app: configures monitoring
- Streamlit app → App configuration: reads and saves
- Streamlit app → Frame capture: starts capture
- Frame capture → Webcam: reads frames
- Streamlit app → YOLO detection: runs inference
- YOLO weights → YOLO detection: loads weights
- YOLO detection → Person detection: provides detections
- Streamlit app → Alert handling: handles detections
- Alert handling → Session alert history: records alerts
- Alert handling → Screenshot evidence: saves evidence
- Alert handling → System beep: starts beep
- Streamlit app → Live dashboard: updates live view
- Streamlit app → Alert log and gallery: refreshes displays
- Session alert history → Alert log and gallery: supplies alert entries