OAK-D, an AI vision kit designed and produced by the OpenCV team, features a built-in Intel® Movidius™ Myriad™ X image processor. Despite its small size, OAK-D integrates a 4K RGB stereo depth camera, IMU, and a high-performance AI processing chip to perform binocular depth visual calculations and neural network inferences. The inertial navigation sensor is integrated into a single camera, allowing users to achieve binocular vision measurement positioning, artificial intelligence neural network acceleration, and real-time 4K H.265 30fps streaming with low power consumption of 2.5W. It caters to the needs of users in smart driving, smart transportation, smart security, robotics, educational competitions, and more.
OAK-D-PoE is based on OAK-D and features a PoE (Power over Ethernet) power supply circuit that allows a single Cat5e (or higher) Ethernet cable to provide power and deliver 1,000 Mbps (1 Gbps) full-duplex connections to devices. With an IP67 protection class housing, it is suitable for use in various environments.
OAK-D-Lite is the most cost-effective product in the OAK USB series. While lacking IMU, its performance is comparable to OAK-D, but at a lower price point. It combines depth sensing, object detection (neural inference), and object tracking, helping you achieve these functions with a simple and user-friendly Python API. OAK-D-Lite includes three built-in cameras (one 4K/30fps RGB camera and two monochrome stereo cameras) and a USB3.0 Type-C interface, allowing you to use it on a regular computer, Raspberry Pi, or other popular embedded host devices via the USB interface.
OAK-D-Pro is an upgraded version of OAK-D, featuring structured light with an IR laser dot projector (for active depth perception) and IR illuminated LEDs (for "night vision"). It is smaller, lighter, and more sensitive than OAK-D. With a built-in high-performance Myriad X VPU, it adopts structured light to improve active binocular vision technology and positioning accuracy to sub-millimeter levels. It meets the needs for applications such as automated material handling robots, surface defect localization, identification, and calibration, enhancing the robot's sensing capabilities.
OAK-D-S2 is a more compact version of ODK-D. It offers the same functions, performance, and lenses as OAK-D but is smaller and lighter. It is suitable for scenes with limited space and a requirement for quality. Compared to OAK-D, OAK-D-S2 removes the 5V power port from the structure and has a wider depth measurement range compared to OAK-D.