cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
deepg799
Explorer
Explorer
728 Views
Registered: ‎01-20-2019

OpenCV - Background subtraction Operation Using Xilinx Vision library

Currently, I am using the createBackgroundSubtractorMOG2() OpenCV function which is running on CPU side.

Is there any support for the similar or same type of function in Xilinx Vision library?

0 Kudos
3 Replies
aoifem
Moderator
Moderator
644 Views
Registered: ‎11-21-2018

Hi @deepg799 

you could try the colour detect function: 

https://xilinx.github.io/Vitis_Libraries/vision/2020.1/design-examples.html?highlight=background

 

There is also a Canny Edge detection example here: 

https://xilinx.github.io/Vitis_Libraries/vision/2020.1/getting-started-with-vitis-vision.html?highlight=background

 

 

 

Aoife
Product Application Engineer - Xilinx Technical Support EMEA


**~ Got a minute? Answer our Vitis HLS survey here! ~**

**~ Don't forget to reply, give kudos, and accept as solution.~**
deepg799
Explorer
Explorer
557 Views
Registered: ‎01-20-2019

Hi @aoifem 

Thank you for your valuable inputs.

I would like to accelerate the OpenCV function like resizing and others on the FPGA by using the Xilinx Vitis-AI Vision library.

My requirement is just to optimize the inference by porting the OpenCV functions on the FPGA side and I am using the Vitis-AI flow for the inference.

It is very new to me So Could you please provide your input on the below queries:

1 - As I can see the different libraries in the below link. such as L1, L2, and so on, which could be suitable for my requirement?

     https://github.com/Xilinx/Vitis_Libraries/tree/master/vision

2 - Is there any step by step tutorial available from the Xilinx side to port these example design on the FPGA side?

3- Can you please provide brief difference information between L1 and L2 library flow.

Thank you in advance.

Tags (1)
0 Kudos
aoifem
Moderator
Moderator
428 Views
Registered: ‎11-21-2018

Hi @deepg799 

I am not an expert on the libraries, but I think you might have them a little confused. There is the Vitis Vision libraries (https://github.com/Xilinx/Vitis_Libraries/tree/master/vision) and the Vitis AI libraries (https://github.com/Xilinx/Vitis-AI). I don't think there are any Vitis-AI Vision libraries. I am not sure which one you need for your application but I've included some information below: 

 

 

Vitis Vision: 

Kernels optimized for Xilinx FPGAs and SoCs, based on the OpenCV computer vision library. The libraries are divided into the following sections: 

aoifem_0-1609777806581.png

 

L1 uses the HLS flow. HLS is a software tool which synthesizes code written in C and C++ into RTL IP blocks. More information on this is here: 

https://www.xilinx.com/html_docs/xilinx2020_2/vitis_doc/introductionvitishls.html

L2 uses the Vitis flow. Vitis is a software platform which enables the development of embedded software and accelerated applications on heterogeneous Xilinx platforms including FPGAs, SoCs, and Versal ACAPs. https://www.xilinx.com/products/design-tools/vitis/vitis-platform.html

I think the other libraries are explained well in the screenshot above and in the link below, but please let me know if you need more clarification. 

For tutorials on how to use them, please use the following link: https://xilinx.github.io/Vitis_Libraries/vision/2020.1/overview.html#

 

Vitis AI Libraries: 

Xilinx’s development stack for AI inference on Xilinx hardware platforms, including both edge devices and Alveo cards. It consists of optimized IP, tools, libraries, models, and example designs for AI. 

The user guide can be found here, and explains who needs it, and how to use it: 

https://www.xilinx.com/support/documentation/sw_manuals/vitis_ai/1_1/ug1354-xilinx-ai-sdk.pdf

https://www.xilinx.com/support/documentation/sw_manuals/vitis_ai/1_3/ug1414-vitis-ai.pdf

There are a number of tutorials here as well: 

https://github.com/Xilinx/Vitis-AI-Tutorials

 

 

Aoife
Product Application Engineer - Xilinx Technical Support EMEA


**~ Got a minute? Answer our Vitis HLS survey here! ~**

**~ Don't forget to reply, give kudos, and accept as solution.~**