Performant embedded machine vision on low cost, low power devices

Performant embedded machine vision on low cost, low power devices

To enable machine vision technologies to run in real time, powerful PC’s or GPU’s are needed to extract the desired data in a split of a second. But for image enabled mobile devices we do not want a massive computer draining the battery of e.g. our AVG, nor do we want a ventilator buzzing next to our ear when using a cobot. And what to think about energy and size requirements of face recognition based entry control systems, medical devices or wearables, which will be the next area image based AI will conquer.

This presentation will give you insight in the techniques that are being used by DPControl to realise high performance machine vision processing on a small and economical package. Applications where small FPGAs and the Myriad VPU are being developed and deployed will be shown and explained in detail.