Three car park functions: bay occupancy, in/out counting, plate recognition.
Traditional ultrasonic parking sensors require drilling and dedicated cabling above every single stall, inflating installation costs; streaming raw video to a central server drains LAN bandwidth and risks site-wide outage on server failure.
Deploy all-in-one edge AI cameras where a single unit monitors 4–6 bays from an oblique angle, determining occupancy entirely inside the camera and reporting lightweight state updates without external servers.
Core Advantages
Traditional inductive loops require cutting entrance pavement and suffer heavy mechanical wear, while manual logs or photocell sensors easily miscount pedestrians, leading to accumulated lot tally errors.
Feed existing RTSP camera streams into an edge AI host, draw virtual detection tripwires, and run tracking algorithms to separate entries from exits in milliseconds, instantly pushing real-time remaining counts to display signs.
Core Advantages
Cloud-dependent license plate recognition suffers from network latency and blocks gates whenever connectivity drops, creating vehicle congestion; proprietary integrated gate cameras carry high replacement and maintenance costs.
Edge computing hosts pull multi-channel high-definition RTSP streams, recognize license plates and vehicle types locally, match against on-premise whitelist databases, and trigger barriers via on-board relays or RS485.
Core Advantages
A single view is judged on the camera, multiple feeds go to an edge host; wet and dusty sites use sealed or fanless models. Channel figures follow the Vision & Video AI technology page.
Not sure which to choose? Read the full selection guide →
Which function do you need in the car park? Each one maps to a reference design.
Single-spot bay detection is handled on-device by reCamera; multi-stream entrance feeds aggregate to reComputer edge hosts for high-performance inference, directly controlling barrier gates via relay, Wiegand, or network.
Core Advantages
| Image | Product | Core Positioning | AI Performance (INT8) | Channels (YOLOv8m) | Cost |
|---|---|---|---|---|---|
 | J4012 | Jetson Orin NX 16G | 100 TOPS | ~7 ch | ~$900–1800 |
 | J3011 | Jetson Orin Nano 8G | 40 TOPS | ~4 ch | ~$600–1000 |
 | RK3588 | Rockchip RK3588 | 6 TOPS | ~2 ch | ~$280–360 |
 | RK3576 | Rockchip RK3576 | 6 TOPS | ~1 ch | ~$160–210 |
 | reCamera | All-in-One AI Camera | 1 TOPS | 1 ch | ~$80–120 |
reCamera 2002 is preferred for standalone bay monitoring; 8–16 IP camera feeds at gates aggregate to reComputer edge hosts (e.g. J4012 / J3011) for license plate recognition and vehicle counting.
Core Advantages
| Image | Product | Cooling Method | Suitable Environment | Operating Temperature | Power Supply |
|---|---|---|---|---|---|
 | reComputer Industrial J4012 | Passive (Fanless) Industrial | Dusty/Workshop | -20~60°C | DC 12V~24V |
 | reComputer Rugged J4012 | IP66 waterproof sealed enclosure | Outdoor/Wet/Rain | -20~60°C | Direct use in 48V systems |
 | reComputer Industrial J3011 | Passive (Fanless) Industrial | Dusty/Workshop | -20~60°C | DC 12V~24V |
 | reComputer Rugged J3011 | IP66 waterproof sealed enclosure | Outdoor/Wet/Rain | -20~60°C | Direct use in 48V systems |
 | reComputer Industrial R2235 | Passive (Fanless) | Dusty/Workshop | -20~50°C | DC 9-36V Wide |
Outdoor poles and exposed gates use IP66 waterproof sealed Rugged models; dusty rooms use fanless Industrial models; DC power sites use models supporting DC 9–36V wide input.
Core Advantages
No. The bay detection model runs locally on the reCamera 2002, monitoring 4–6 bays per unit. The camera also serves the Node-RED dashboard directly, accessible via browser.
Yes. Entrance IP cameras connect via RTSP to a reComputer edge host (such as J4012 / J3011) to run LPR. Matching against a local whitelist can trigger the barrier gate within seconds via relay IO, Wiegand, or RS485.
Yes. Entrance cameras connect to a reComputer edge host running tripwire detection and bidirectional counting, calculating real-time vehicle flow and vacant spaces, pushing updates via MQTT to LED signs or management systems.
Yes. Any IP camera supporting RTSP can connect directly to a reComputer edge host over LAN for video analytics without re-cabling or replacing cameras. Supported channels are listed in the Compute & Channels table above.
Yes. The Seeed devices used in this solution can be customized with your own logo, enclosure, packaging and pre-loaded firmware, and existing models can be adapted to add or drop interfaces and features. See Customization Service for scope and process, or .
Docs
