/ImageAlgorithmAccelerationByHLS

Use HLS to accelerate image algorithms

Primary LanguageC++MIT LicenseMIT

ImageAlgorithmAccelerationByHLS

Introduction

Image enhancement-gamma transformation

Designed by woodfan

Origin Picture:

Processed picture:

Figure 3 Processing plus dma transmission takes 34ms, which takes 653ms on average to process the same image under vs2015 where the hardware platform is i7-9750H 2.6GHz with 16GB RAM. The performance difference is nearly 20 times.

Vivado Block Design:

FPGA Platform: PYNQ-Z2

Pseudo-color grayscale

Designed by woodfan

Not Work

rgb2gray

Designed by Fu Guohang

Not Work

Thanks to xilinx for organizing this event, and thank my teammate, Fu Guohang, who is not very familiar with hardware and has a lot of difficulties in learning but persisted. Thanks to my teammate, Yang Moyuan, who is still working during the day and insists on providing us with a lot of useful learning materials at night. Thanks to my teaching assistant colleague Li Yang, who has taken care of me a lot in class management

2020年新工科联盟-Xilinx暑期学校(Summer School)项目
2020 New Engineering Alliance -Xilinx Summer School Project

To Be Continued