Banana Pi BPI-OM7 is an AI 3D depth camera that pairs the Banana Pi BPI-M7’s low-profile Rockchip RK3588 SBC with an ORBBEC Gemini 2 depth camera, aimed at 3D vision, robotics, edge AI, and spatial perception applications.
This setup includes 8GB of RAM and a 64GB eMMC flash by default, features HDMI and USB-C video outputs, dual 2.5GbE networking, and several USB ports. It’s conveniently mounted on a tripod.
Specifications include:
– SoC: Rockchip RK3588 octa-core processor (4x Cortex-A76 cores @ 2.4 GHz, 4x Cortex-A55 cores @ 1.8 GHz)
– GPU: Arm Mali-G610 MP4
– Video Decoder: Supports 8Kp60 H.265, VP9, AVS2, 8Kp30 H.264 AVC/MVC, and more
– Video Encoder: 8Kp30 H.265/H.264
– AI Accelerator: 6 TOPS NPU
– RAM: 8GB (default), 16GB, or 32GB LPDDR4x
– Storage: 32GB, 64GB (default), or 128GB eMMC, M.2 NVMe SSD, microSD slot
– Video Output: HDMI 2.1 up to 8Kp60, USB-C with DisplayPort Alt mode up to 8Kp30
– Camera: ORBBEC Gemini 2 (RGB camera and a 3D depth camera with active stereo IR)
– Audio output via HDMI or USB-C, internal speaker connector
– Networking: Dual 2.5 GbE RJ45, WiFi 6, and Bluetooth 5.2
– USB Ports: USB 3.0, USB 2.0, USB 3.1 Type-C with DP 1.4 Alt. mode
– Expansion: M.2 Key-M 2280 (PCIe 3.0 x4) socket, 40-pin Raspberry Pi-compatible header
– Misc: 2x LEDs, RTC battery (TBC)
– Power Supply: Via USB-C PD port
Documentation highlights the specs separately for the SBC and camera, but the Getting Started Guide is crucial, showing how to use Ubuntu 24.04 OS with Docker to install the Orbbec SDK, the RKNPU/RKNN toolkit, and run various demos using the combo.
It also supports the Orbbec Viewer. You can find more resources on GitHub in the BPI-OM7-orbbec_reconstruction repo. The Banana Pi BPI-OM7 AI 3D camera is available on AliExpress for $739.36 plus shipping and taxes, which is compared to other similar products like the NVIDIA Jetson Nano-based Orbbec Femto Mega.
Support CNX Software by donations or purchasing goods through affiliate links.
