Run face detection algorithm with hardware acceleration on the Zedboard.
The code is modified from https://github.com/cornell-zhang/facedetect-fpga.
Develope Environment: Windows SDSoC 2016.2
Simply pull all of the things under sdcard_image folder to the SDCard.
-
Open SDSoc -> create new SDSoC project -> enter project name, set Target Platform as zed. -> choose Empty Application and click Finish.
-
Right click the 'src' folder under the newly create project -> Choose Import -> File System -> Browse to this project's src folder and import all of the file into the project.
-
Right click the icon 'left' next to haar.cpp and it will show the funcitons in the haar.cpp -> left click detectFaces and toggle it as HW function.
-
Click the navigation bar on the top and choose Build Project.
-
After building finished, it will generate the sd_card folder under the SDDebug.Pull all of the things under SDCard_image to your SDCard. And boot the Zedboard form the SDCard.
cd /mnt
./face_detect_baseline_plus.elf 192x144_pgmfile(there are five test pgm files to use)
The program will create new file to store the result.