Starter kit for on-device computer vision

Our solution

We have created a Starter Kit that enables faster development and deployment of new, innovative computer vision systems. It’s powered by:

  • Raspberry Pi 3,
  • low power high performance Intel® Movidius™ Myriad™ 2 Vision Processing Unit (VPU)
  • Basler’s camera of the following configuration: camera daA1600-60uc (2MP, 1600 px x 1200 px, 60pfs) + lens Evetar M12B0416IR F1.6 f4mm 1/2″.
  • byteLAKE’s Computer Vision asynchronous model to enable real-time on-device objects classification.

Additional features

  • designed as a low-power solution, especially for devices where battery life is critical,
  • ensures high performance with the use of hardware accelerators dedicated for embedded imaging and vision applications,
  • possible to embed the solution into small form factors.

It can be applied across wide selection of industries and can add intelligence to your i.e. security cameras, drones, industrial machine vision equipment, and more. Have a look at our video demonstrating an example application: detection of groceries using a camera.

AI on Edge brings a completely new quality to the IoT world in many ways. Have a look at our blog post here, where we summarized our thoughts about on-device vs. in-the-Cloud (or Data Center) deployment strategies for AI solutions. We also benchmarked two of the most commonly used AI hardware accelerators.

on-device computer vision

See the video here.