The engineers and product designers at [moovel lab] have created the Open Data Cam – an AI camera platform that can identify and count objects as they move through its field of view – along with an open source guide for making your own.
Step one: get out your ruler and utility knife. In this world of ubiquitous 3D-printers they’ve taken a decidedly low-tech approach to the project’s enclosure: a cut, folded, and zip-tied plastic box, with a cardboard frame inside to hold the electronic bits. It’s “splash proof” and certainly cheap to make, but we’re a little worried about cooling and physical protection for the electronics inside, as they’re not exactly cheap and rugged components.
So what’s inside? An Nvidia Jetson TX2 board, a LiPo battery with some charging circuitry, and a standard webcam. The special sauce, however, is the software, which is available on GitHub. [Moovel lab]’s engineers have put together a nice-looking wifi-accessible mobile UI for marking the areas where you’d like the software to identify and tally objects. The actual object detection and identification tasks are performed by the speedy YOLO neural network, a task the Nvidia board’s GPU is of course well suited for.
As the Open Data Cam’s unblinking glass eye gazes upon our urban environments, it will log its observations in an ancient and mysterious language: CSV. It’s up to you, human, to interpret this information and use it for good.
A summary video and build time lapse are embedded after the break.
Alex, I’ll take Arti’s Anal for $600.
Lol. Connery is the man!
For what one Jetson TX2 module costs, I can pay a college student to stand in the rain and count cars for a week.
+1 :D
And make her count cars? What a waste of resources.
Hmmmmmnnn …… I’d say this article is misleading in such as fantastical manner as to conclusion that it’s a joke. This is becuase setting up Jetson TX2 is not a ‘trivial’ matter like programming an Arduino. Be warned !!!!!! BTW – don’t forget to take off the lens cap on your Jetson onboard camera – the results are sooooooo much better!
At least the enclosure is cheap.
“The engineers and product designers at [moovel lab] have created the Open Data Cam – an AI camera platform that can identify and count objects as they move through its field of view – along with an open source guide for making your own.”
Which is what I predicted placing the smarts camera-side would be. Now all they have to figure out is making it all cheap enough for everyone.
Gotta love CSV.
Wonder why recognition part in the video is accelerated ? Using pause button, it’s easy to see a lot of cars missing, peoples labeled as car and even a group of cars labeled as aeroplan ! The rest of the video is almost pointless… I suspect that almost every mid range new smartphone could do that low level of categorization quality.
So… it’s just a camera, and a thousand bucks worth of off-the-shelf standard Jetson board, and stock-standard YOLO.
You could at least *try* and make something that is novel, or takes effort, and isn’t massively bloated in power and cost.
Do it on a Raspberry Pi with a Movidius stick and knock off 80 percent of the price, with a little bit of a hit in frame rate – even that is still just plugging together stock standard parts.
Yes, raspberry pi at 1 Hz frame rate would probably be fine for this application. This kind of system would indeed be better exploited elsewhere.
the google aiy vision kit has a bonnet for the pi0w that is based on the Movidius MA2450 chip so camera, pi, and VPU for under a hundred bucks, and the next AIY project the “edge TPU” looks very interesting depending on what they can hit for a price point.