New in the SevinHub suite: RadarHub watches any video feed, detects every vehicle, measures its real speed in km/h, counts traffic by type, and flags speeders with photo evidence. It runs entirely on your own hardware, and you can try it in your browser right now.
Some projects are business decisions. This one is a childhood dream.
You have probably seen the viral story: a student spent 20 dollars on AI credits, built a video speed radar, and sold it for a life-changing amount. When I saw it, I realized two things. First, the technology behind it is real and well understood. Second, there was no reason I could not build my own version, properly, and make it a product anyone can own.
So I did. Meet RadarHub.
Point it at any road camera and it turns raw video into traffic intelligence:
I built a live demo that runs the AI entirely on your device: your camera feed never leaves your phone or laptop. Point it at a street and watch it detect, track and count.
Open the RadarHub live demo (works on mobile, camera permission required).
RadarHub accepts three input types: a video file, a live RTSP/IP camera stream (which covers virtually every CCTV camera made in the last decade), or a webcam, including OBS Virtual Camera. Calibration is point-and-click: you mark 4 points on the road, tell it the real size in meters, and from then on speeds are measured on the ground plane.
This was non-negotiable for me. RadarHub runs 100% on your own hardware. There is no cloud processing, no per-camera monthly fee, and no footage ever leaves your machine. For anything traffic-related, that privacy model matters.
One honest note: official speeding fines usually require type-approved devices. RadarHub is a measurement and analytics tool, built for studies, private sites, monitoring and evidence gathering.
Like everything on SevinHub: one payment, yours forever. 99 EUR for the Standard license with full Python source code, or 299 EUR Extended with white-label and client-project rights. Updates included.
Get RadarHub on the marketplace · Feature overview · Documentation
A dream from when I was a kid is now a product you can download. That is a good day.