Frigate, an open source NVR that uses AI to recognize and analyze moving objects captured on camera in real time
Advances in smart home technology and compatible devices have made it cheap and easy to install surveillance cameras in your home. ' Frigate ' is an open source
Frigate NVR
https://frigate.video/
The Frigate demo is available on the following site, and you can experience what kind of NVR it is.
Frigate
https://demo.frigate.video/
When you access the Frigate demo site, it looks like this. Video of the crosswalk taken by a surveillance camera is displayed.
If you select 'Event' from the left, the moment the automatically detected moving object was photographed will be displayed.
'Exports' allows you to filter by camera type and date and time, and play and download created events.
In 'Storage', the storage capacity for saving video, memory usage rate, and network bandwidth for distribution were displayed.
If you look at 'System' you will see CPU usage and GPU usage.
Click 'Config' to change settings.
'Log' allows you to view system logs.
Traditional NVRs have simple motion detection that can require hours of fine-tuning to reduce false positive rates. By offloading object detection to the TPU of Google's local AI platform ``Google Coral'', Frigate enables advanced analysis even on low-spec hardware, allowing users to detect objects such as people, cars, or other Determine whether it is an object. Frigate can perform over 100 object detections per second.
Additionally, Frigate can be integrated directly into the home assistant's media browser, making it possible to integrate object detection into a variety of home assistants. Frigate is developed as open source, and the source code is available in the GitHub repository below.
GitHub - blakeblackshear/frigate: NVR with realtime local object detection for IP cameras
https://github.com/blakeblackshear/frigate
Related Posts:
in Review, Software, Web Service, Posted by log1i_yk