PYNQ: PYTHON PRODUCTIVITY

Flow to merge PYNQ DPU and xfOpenCV?

I have been redirected to post my question here from the Xilinx forums.

I am looking for a tutorial for adding IP cores to a PYNQ base image.

Right now I have gotten PYNQ with a webcam and the YoloV3 model working according to this guide:

I have added some filters in OpenCV in software too but now I want to accelerate it in hardware using xfOpenCV.

I am using an Ultra96-V2 and I am successful in building the base overlay.

And right now I am confused how to add the DPU and xfOpenCV IP cores to the overlay in Vivado.

Can someone guide me through the steps on compilation and integration of the IP cores? I also understand that PYNQ overlays need a slightly different process.

You will have to build your customized Vitis platform, then use Vitis AI to integrate DPU onto that platform. For your customized Vitis platform, you can add your IP when building it.