How to use PYNQ-Z2 for image algorithm acceleration

Hello everyone!
Now, I plan to use PYNQ-Z2 for image capture. I have learned how to use python for image acquisition when the camera is connected to the USB host port. But this is done entirely on the PS side.
Now I plan to accelerate the image algorithm, I plan to use HLS to write an IP CORE,
My idea is to use PYNQ’s base overlay, use the “source base.tcl” tcl command to rebuild the vivado project, and then put the written IP core between the vdma and hdmi out IP cores. Then repackage it into a new overlay, and then call it on the PS side
summary of a problem:

  1. Is my idea correct?
  2. Does the pynq repository on GitHub have to be rebuilt with vivado 2020.1? I tried to use 2018.3 before and found it was not successful.

You can find an example here that may be of use:
You don’t need to build the Vivado PYNQ base project, but building it may help you learn more about Vivado and Zynq designs.

You should match the correct version of Vivado with the version of PYNQ you are using, and if you are rebuilding an existing project, you should use the same version that was used to create it.


Hi Cathal!
Thanks for your reply, it helped me a lot!