Viseron
NVR software that records IP cameras and detects motion, objects, faces and license plates locally, with a web interface for setup and playback.
v3.7.0 · Stable•MIT•Since Aug 2020
Overall rating
activity, maturity, stars
7.3
/ 10
GitHub stars
3.7k
Contributors
28
Last commit
15h ago
Project age
6 years
About
Viseron is a network video recorder (NVR) with AI computer vision, built for anyone who wants to keep an eye on a home, office or other space. It captures video from IP cameras and analyzes it entirely locally, with no cloud involved. Object detection, motion detection and face recognition are the headline features.
Its functionality is modular and enabled through components, which you configure in a built in web interface after spinning up a Docker container.
Key features
- Object detection using frameworks such as YOLO, Darknet and TensorFlow
- Motion detection
- Face recognition
- License plate recognition
- Hardware acceleration with Google Coral EdgeTPU and CUDA
- RTSP and IP camera support
- Configuration through a built in web interface
- Modular component based architecture
Ways to run it
- Docker container