基于Zynq和改进ViBe算法的运动目标检测系统设计  

Design of Motion Target Detection System Based on Zynq and Improved ViBe Algorithm

作　　者：武靖 姜冬瑞 赵富成 张会新[1] WU Jing;JIANG Dongrui;ZHAO Fucheng;ZHANG Huixin(Key Laboratory of Micro/nano Devices and Systems,Ministry of Education,North University of China,Taiyuan 030051)

机构地区：[1]中北大学微纳器件与系统教育部重点实验室,太原030051

出　　处：《舰船电子工程》2024年第12期162-166,共5页Ship Electronic Engineering

基　　金：国家自然科学基金项目(编号:51705475)资助。

摘　　要：针对原Vi Be算法存在的“鬼影”和运动目标静止时被更新为背景模型的问题提出改进ViBe算法;并针对当前运动目标检测系统平台大,功耗高等问题,设计了一种基于Zynq平台的运动目标检测系统,使用Vivado HLS进行接口为AXI4-Stream的改进ViBe算法IP核的开发。在Zynq7020平台,使用PL端作为算法的实现单元,PS端作为控制核心,充分发挥异构So C的架构优势。实验结果表明:改进ViBe算法解决了“鬼影”问题和运动目标静止时被更新为背景的问题,系统实时性较高,在保证对运动目标检测准确性的基础上降低了系统功耗,且体积小。The improved ViBe algorithm is proposed to address the"ghosting"problem of the original ViBe algorithm and the problem that the moving target is updated to the background model when it is stationary.And to address the problems of large plat⁃form and high power consumption of the current moving target detection system,a moving target detection system based on the Zynq platform is designed,and the IP core of the improved ViBe algorithm is developed using the Vivado HLS.The development of the im⁃proved ViBe algorithm IP core with AXI4-Stream interface is carried out.In the Zynq7020 platform,the PL side is used as the im⁃plementation unit of the algorithm,and the PS side is used as the control core,which gives full play to the architectural advantages of heterogeneous SoCs.The experimental results show that the improved ViBe algorithm solves the"ghosting"problem and the prob⁃lem of updating the moving target to the background when it is stationary,and the system has high real-time performance,it reduc⁃es the power consumption of the system on the basis of guaranteeing the accuracy of detecting the moving target,and has a small size.

关 键 词：改进ViBe算法 运动目标检测 Zynq Vivado HLS VDMA 

分 类 号：TN919.5[电子电信—通信与信息系统]