Using OAK Cameras on a Vehicle for City Object Detection

nuradiiil

Hello everyone,

I’m currently working on a vehicle-based system that can detect and classify various urban objects (cars, pedestrians, road signs, billboards, etc.) while the vehicle is in motion.

I’m considering using Luxonis OAK cameras for this project and would like to ask the community (and the Luxonis team):

What components or accessories are required to set up such a system on a vehicle?

Is the OAK-4D the best option for mobile AI and depth applications?

Which pre-trained AI model would you recommend for this kind of task?

And what is the best way to integrate GPS data with detection results for mapping purposes?

The main idea is to build a system that detects objects and records their coordinates in real time for urban environment analysis, adding these detections to our geoinformation platform.

Any advice, recommendations, or examples of similar projects would be greatly appreciated!

jakaskerl

Hi nuradiiil
For OAK4, you only need a PC for initial setup. For powering, POE+.

nuradiiil Is the OAK-4D the best option for mobile AI and depth applications?

If you want the device to work standalone with high AI workload + depth, then yes.

nuradiiil Which pre-trained AI model would you recommend for this kind of task?

Probably some widely trained yolo. Feel free to check our pretrained models and pick one that you like: https://hub.luxonis.com/ai/models/trending?search=

nuradiiil And what is the best way to integrate GPS data with detection results for mapping purposes?

It should generally be possible to attach a USB-C GPS dongle, but some modifications to driver will likely need be needed.

Thanks,
Jaka