OAK-D CM4 PoE, OpenCV AI Machine Vision Kit, Depth Measuring, Image Recognition

Overview :

The OAK-D CM4 PoE is an upgraded version of OAK-D-CM4, as PoE is used both for power delivery and communication for the Rasbperry Pi Compute Module 4, plus it comes in a IP67 rated enclosure. The enclosure is about the size as the OAK-D-PoE. The color camera is also close to the right camera, so RGB-depth alignment isn’t needed in most cases, which reduces system load.The OAK-D CM4 PoE combines a host (RPi CM4) and the OAK-SoM-Pro to allow for a fully standalone and integrated solution for real-time Spatial AI.

Hardware specifications :

This OAK camera uses Power-over-Ethernet (PoE) for communication and power. It offers full 802.3af, Class 3 PoE compliance with 1000BASE-T speeds (1 Gbps) and has a micro SD (uSD) card connector. A PoE injector/switch is required to power the device.

Myriad X inside :

This OAK device has on-board Myriad X VPU 

4 TOPS of processing power (1.4 TOPS for AI)
​Run any AI model, even custom architectured/built ones - models need to be converted.
Encoding: H.264, H.265, MJPEG - 4K/30FPS, 1080P/60FPS
Computer vision: warp/dewarp, resize, crop via ImageManip node, edge detection, feature tracking. You can also run custom CV functions
Stereo depth perception with filtering, post-processing, RGB-depth alignment, and high configurability
Object tracking: 2D and 3D tracking with ObjectTracker node.

Dimensions and Weight :

Weight: 318g

Depth perception :

This OAK camera has a baseline of 9.0cm - the distance between left and right stereo camera. Minimal and maximal depth perception depends on camera FOV, resolution, and baseline.

• Min perceivable distance: ~20cm (400P, extended), ~40cm (400P OR 800P, extended), ~80cm (800P)

• Max perceivable distance: ~35 meters

Extended means that StereoDepth node has Extended disparity mode enabled.

Links related to OAK-D CM4 PoE : Datasheet