OpenCV on Beaglebone with image recognition

Published on March 13, 22

Beaglebone
Beaglebone

Technologies used

  1. Synpse Cloud for hosting and running applications
  2. Beaglebone AI Beaglebone AI is hardware that can be used to build image recognition applications.

Beaglebone with Synpse

Beaglebone boards comes with specialized OS’es with built in development tools. Downside is that these devices now ships without GUI support. We have user guide how to prepare Beaglebone and install Synpse agent.

Once you have it up and running, we can use synpse CLI to access remote device:

1
synpse device proxy beaglebone --port 8080:80

You can access Cloud9 IDE from your device:

Cloud9
Cloud9

Once you have access to device, you can deploy test application to it. Connect web-cam and run classification.tidl.c from /cloud9/Beaglebone/AI/tidl/ directory.

Image recognition
Image recognition

You can access the feed via second port, run command bellow in the second terminal:

1
synpse device proxy beaglebone --port 8090:8090

An check if it works:

Coffee mug!
Coffee mug!

Wrapping up

Beaglebone is very nice platform to build image recognition applications. With Synpse this can be deployed and managed remotely.