Steps to Prototype a Streaming 4x Application on a Small Embedded Solution

Embedded vision components are becoming increasingly popular and are being integrated into a wide range of applications. What these applications have in common is the need to pack more functionality into a small space. It is often beneficial when these systems make decisions at the edge.

To make such a system possible, including the ability to rapidly prototype, Teledyne FLIR introduced the Quartet™ TX2 embedded solution . This custom carrier board easily integrates up to 4 full-bandwidth USB3 machine vision cameras.

To highlight what is possible with Quartet pre-installed with the Spinnaker SDK, we walked through the steps we took to make an ITS (international transportation system) inspired prototype that can run four similar applications simultaneously, three of which use deep learning.

The details are as follows:

• Application 1: License plate recognition through deep learning

• Application 2: Classifying vehicle models through deep learning

• Application 3: Vehicle color classification using deep learning

• Application 4: Viewing forward through the windshield (eliminating reflections and glare)