🌐 Languages: [English] | [한국어]
(An emergency situation detection & reporting system leveraging existing CCTV infrastructure)
CCTV-EDR is a system that automatically detects emergency situations using CCTV infrastructure and immediately connects detection to reporting and real-time monitoring. It uses an object detection-based AI model on CCTV footage to automatically identify situations where a person collapses, faints, or loses consciousness. After detection, it enables rapid emergency response through a real-time alert and reporting workflow designed to minimize latency. Operators can also monitor the CCTV stream where the emergency was detected in real time.
- Competition: 2021 WINNER LINC+ Capstone Design Untact Competition
- Period: 2021.09 - 2021.12
- Host: Ministry of Education, LINC
- Award: 🥈 2nd Prize
- ©️ No. C-2021-057286 호
- YOLOv5
- NodeJS
- Nginx
- MongoDB
- Docker
- Real-time object detection on CCTV footage
- Real-time server notifications for detection results
- WebSocket-based real-time monitoring
demo.mp4
- Stabilize and harden the server and database for reliable data processing
- Improve object detection model performance and reduce false positives/false negatives
The code in this repository is released under the MIT License.