Based on a SMARC computer-on-module with an i.MX 8M Plus processor, it allows users to get acquainted with NXP’s neural processing unit (NPU).
Delivering up to 2.3Top/s for deep learning based artificial intelligence, it can run inference engines and libraries such as Arm Neural Network and TensorFlow Lite, according to Congatec. It also integrates with Basler embedded vision software for developing AI accelerated systems – Basler pylon camera software has a unified SDK for BCON for MIPI, USB3 vision and GigE vision cameras, and enables camera access from source code, GUI or 3rd party software.
“The pylon viewer is perfect for camera evaluation,” according to Congatec. “Engineers get access to AI supported machine vision features such as triggering, individual image capture, and camera configuration options plus customised inference algorithms on the basis of the Arm NN and TensorFlow Lite ecosystem.”
In the kit, a passively-cooled conga-SMX8-Plus board is included for its quad Cortex-A53 cores, Cortex-M72 controller and NXP NPU. A 3.5 inch conga-SMC1/SMARC-ARM carrier board directly connects the included 13Mpixel Basler ‘daA4200-30mci BCON for MIPI camera’ (with F1.8 f4mm lens) via MIPI CSI-2.0. USB and GigE vision cameras are also supported.
A bootable SD card includes a pre-configured boot loader, Yocto OS image, matching board support packages and Basler embedded vision software for AI inference training on captured images and video sequences.
Congatec provided a list applications that could be developed with the kit: automated checkout terminal for retail, building safety equipment, in-vehicle vision for navigation, surveillance systems in busses, user-identification for industrial computer interfaces, gesture-based machine operation, vision supported robots and industrial inspection systems.
Congatec will be presenting its i.MX 8 processor ecosystem at virtual NXP Technology Days – 8 – 30 June.