OpenCV Launches AI Kit, OAK, Family of Open-Hardware Myriad X-Based Computer Vision Boards

OpenCV, one of the biggest open source computer vision projects around, has launched a crowdfunding campaign for an open hardware kit to go alongside its software: the OpenCV AI Kit, or OAK.

"OAK is a modular, open source ecosystem composed of MIT-licensed hardware, software, and AI training," explains OpenCV's Brandon Gilles, "that allows you to embed the super-power of spatial AI plus accelerated computer vision functions into your product. OAK provides in a single, cohesive solution what would otherwise require cobbling together disparate hardware and software components."

"OAK consists of the OAK API software and two different types of hardware: OAK-1 and OAK-D. They are tiny artificial intelligence (AI) and computer vision (CV) powerhouses, with OAK-D providing spatial AI leveraging stereo depth in addition to the 4K/30 12MP camera that both models share. They are also both absurdly easy to use. Up and running in under 30 seconds, OAK-1 and OAK-D allow anyone to access this power: hobbyists, researchers, and professionals alike. Once you're done tinkering, OAK's modular, FCC/CE-approved, open source hardware ecosystem affords direct integration into your products."

Which OAK board you need will depend on your use cases. Both include neural inference engines, warp and dewarp functionality, up-to-20-objects tracking, structured navigation via AprilTags, H.264 and H.265 hardware encoding, JPEG encoding for 12 megapixel stills, motion JPEG (MJPEG) encoding, feature tracking, motion estimation, and edge detection with Harris filtering. The OAK-1 adds automatic motion-based lossless zoom capabilities, offering up to 12x lossless zoom at 720p down to 1.5x lossless zoom at 4k; the OAK-D offers stereo depth capabilities with object tracking through 3D space and 3D object localization.

Both boards are based on Intel's Myriad X neural network processor, but OpenCV claims its implementation beats the competition not only on price but on capabilities. The boards are compatible with the Raspberry Pi and other Linux-based single-board computers, and can offer enough compute performance to capture 12 megapixel stills and 4k video with near-zero CPU time and at a peak 6W power draw — including that required by the Raspberry Pi host.

Full details of the boards can be found on the Kickstarter campaign page, where the OAK-1 is priced at $99 and the OAK-D at $149 — roughly half their expected retail prices — with shipping due in December this year.