摘要
传统安防通常是由安保人员24 h值班,轮流监察,发现异常问题再进行处理的过程。但人难免会因各种因素导致异常情况发生时不能及时采取有效防控措施的情况出现。对深度学习YOLO算法进行改进,结合嵌入式平台体积小、易部署、成本低等特征,设计出了一个基于嵌入式的行人检测、追踪智能安全防控系统。系统包含行人检测、行人追踪、行人重识别、数据库管理、显示界面5个模块,很好地实现了实时监测、精准确定和追踪目标、发现异常情况及时报警的功能,真正达到了智能安全防控的目的。
Safety prevention and control is the foundation of people’s livelihood, and security issues should not be underestimated.Traditional security is usually a process of 24-hour duty, turntable, and discovered abnormal problems.But people will inevitably have events that cannot be effectively prevention and control measures due to various factors.This paper introduces pedestrian detection and pedestrian tracking based on the deep learning YOLO algorithm for improvement, and combines the characteristics of embedded platforms, easy deployment, low cost, etc.,the intelligent security control system based on embedded downstairs detection and tracking.This system contains 5 modules: pedestrian detection, pedestrian tracking, and pedestrian re-recognition, database management, display interface.It is a good time to achieve time monitoring, accurate identification goals, tracking objectives, and find abnormalities and timely alarms.It is true for intelligent security prevention and control purposes.
作者
连俊芳
LIAN Junfang(Practical Teaching Center,Minnan lnstitute of Technology,Shishi 362700,China)
出处
《技术与市场》
2022年第9期19-22,共4页
Technology and Market
基金
2021年闽南理工学院校级科研项目(21KJX030)。
关键词
行人检测
行人追踪
行人重识别
智能安全防控
pedestrian testing
pedestrian tracking
pedestrian re-recognition
intelligent
