基于PYNQ平台的人脸检测系统设计  

Design of face detection system based on PYNQ platform

作　　者：王洁[1] 李鹏飞 WANG Jie;LI Pengfei(School of Software Technology,Dalian University of Technology,Dalian 116620,China)

机构地区：[1]大连理工大学软件学院,辽宁大连116620

出　　处：《实验室科学》2023年第4期40-43,47,共5页Laboratory Science

基　　金：大连理工大学研究生教改基金(项目编号:JG_2019065)。

摘　　要：ZYNQ平台提供了ARM+FPGA架构用以提高系统执行效率。PYNQ平台在ZYNQ基础上添加了linux+python环境,降低使用异构平台开发计算机视觉算法难度,因此对基于PYNQ的人脸检测系统进行研究。使用vivado工具链设计并实现人脸检测核心算法、使用高层次综合工具将检测算法定制为可复用IP核、添加粘合逻辑将定制的IP核与ZYNQ硬核相连、编写上位机驱动程序控制实现ARM和FPGA数据交互过程、实现数据采集并以视频流输入结果等步骤。此实验案例以软硬件协同设计为指导思想,充分发挥FPGA的并行计算性能和ARM程序设计灵活性,有效提高人脸检测系统执行效率。Zynq platform provides ARM+FPGA architecture to improve system execution efficiency.Linux+Python environment is transplanted on the Zynq by PYNQ which reduces the difficulty of developing computer vision algorithms with heterogeneous platforms.Therefore,the face detection system based on PYNQ is studied.The core algorithm of face detection is designed and implemented with vivado platform.The detection algorithm is customized as a reusable IP core with high-level synthesis tools.The customized IP core is connected with the Zynq hard core by glue logic.The data interaction between ARM and FPGA is controlled by upper computer program.Realizing data acquisition and inputting the results with video stream.This experimental case takes the software and hardware collaborative design as the guiding ideology,giving full play to the parallel computing performance of FPGA and the flexibility of ARM programming and effectively improving the execution efficiency of face detection system.

关 键 词：PYNQ平台 人脸检测 软硬件协同设计 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]