In the vast field of artificial intelligence, computer vision and image recognition is perhaps what most people take interest in.
If you are interested in this field, you must have heard of OpenCV. OpenCV is a popular open source project aimed at real-time computer vision.
The OpenCV project has announced its hardware project: OpenCV Artificial Intelligence Kit (OAK). It is basically a Raspberry Pi like single board computer specially focused on Computer Vision. This project is running a Kickstarter funding campaign.
If you’re someone already working on computer vision, you may have heard of Nvidia Jetson Nano developer kit as one of the Raspberry Pi alternatives tailored for AI projects.
Even though Jetson Nano could be a better option for AI projects, it looks like OpenCV AI kit actually makes it easy to get started with building projects with OpenCV out of the box.
Let’s take a look at some details.
OpenCV AI Kit Overview
OpenCV AI Kit (OAK) is an MIT-licensed open source software and Myriad X-based hardware solution for computer vision by OpenCV (if that wasn’t obvious).
You can train your own neural networks or just get started with tracking and detecting things using the existing neural networks that include mask/no-mask detection, age recognition, face detection, object detection, vehicle detection, and more.
Not to forget, you can always use OpenVINO to deploy your own model using any available dataset.
You will find two variants of the kit — OAK-1 and OAK-D.
The OAK-1 includes a 4K AI camera module and the OAK-D features a three-camera setup, both leveraging the stereo depth sensor.
OAK-1 and OAK-D supports Linux, Windows, and macOS as hosts.
Even though the OAK API software lets you do a lot of things, there are some board-specific features for OAK-1 and OAK-D that you can find on their Kickstarter campaign page.
As an overview, the official announcement sums up what the OpenCV AI Kit is made up of:
- A single-camera 4k @ 60fps hardware module which includes a Myriad X and is a tiny 45 mm x 30 mm.
- A module with a 4k @ 60fps camera and stereo depth cameras which provide spatial 3D tracking capability. It is about the size of a Raspberry Pi.
- A nsoftware library for advanced on-device real-time neural network processing for the OAK boards.
- Both boards can run Deep Learning models for image classification, object detection, segmentation,
human pose estimation, and many more in real time even on low-power hosts like the Raspberry Pi.
OpenCV AI Kit Pricing & Availability
For now, you can’t get your hands on it without being a backer on Kickstarter.
But, it’s good to see that the funding goal of $20,000 was completed in about 20 minutes of launching the campaign or so as they claim.
However, you might want to join in as a backer on Kickstarter because the retail pricing of the AI kit will be twice as what you see right now.
The OAK-1 is priced at $99 and the OAK-D is available for $149, which was just $79 and $129 for early bird backers.
After the Kickstarter campaign ends, you may notice a retail price of $199 and $299 respectively.
Looking at other Myriad-X based options, OpenCV AI Kit looks a lot more promising, cheaper, and easy to get started.
And, considering that they’re a company you can vouch for — it might be a good time to become a backer if you want to get your hands on it.
Wrapping Up
Looking at the pricing and the feature list, it’s definitely a unique offering to what already exists out there.
What do you think about it? Let me know your thoughts on the same.