CASIA OpenIR  > 毕业生  > 硕士学位论文
Abstract基于视频监控的人脸识别技术是融合模式识别、视频图像处理、计算机视觉等多学科的具有挑战性的课题,有效解决了无法快速定位监控视频中可疑人身份问题,由于它提供了一种可靠、方便的身份鉴别途径,因此,具有广泛的应用环境和广阔的应用前景,近年来成为研究的热点。 本文主要研究在大范围视频监控环境下,利用人脸识别技术对监控画面中的人员进行快速身份定位。主要研究内容包括:对出现在监控视频中的人脸进行快速准确的定位,截取人脸照片,并且提取人脸特征;对不同型号的监控摄像机可以快速方便的接入人脸识别系统,并且可以实时监测摄像机的状态,可以快速自动完成摄像机掉线的重连工作;在多路监控视频同时工作环境下,能够有效调度服务器资源进行实时人脸识别;对待目标人脸数据库过大时能够对采集的人脸进行快比对;对课题进行需求分析、概要设计、详细设计等环节,最后采用C#和VC++语言初步实现了一款可在大规模人脸数据以及大范围视频监控环境下进行实时人脸识别的系统—大规模人脸识别监控系统。 该系统有5个子系统构成,分别为:人脸数据采集系统,负责监控摄像机视频解码以及对出现在视频画面的人员进行人脸检测、人脸跟踪、特征提取和人脸截图功能;人脸特征比对系统,主要负责完成人脸识别工作;负载均衡系统,主要完成人脸数据采集系统与人脸特征比对系统的数据转发工作;基于Web的数据库管理系统,主要完成人脸数据的增删改查以及人脸识别日志维护工作;黑名单报警系统,主要为对识别的可疑人员进行实时报警展示工作。    最后通过测试验证了大规模人脸识别监控系统符合课题功能和性能的要求,完成了论文的基本任务,为本文系统在实际中的应用做好准备工作。
Other AbstractFace recognition technology based on video surveillance, is a multidisciplinary challenging subject which is the integration of pattern recognition, video image processing and computer vision. It has a wide range of applications and promising prospects, and has become a research hotspot in recent years, because it provides a highly reliable and convenient identification recognition way. This thesis puts forward a fast way to identify users in a wide range of video surveillance by face recognition technology. The main content of this research includes quickly and accurately locating the face region in the video image, intercepting face photos and extracting facial features. The system is able to be compatible with different types of cameras, real-time monitor the state of the cameras, reconnect with the cameras automatically and quickly when the line drops, schedule server resources effectively and search face s in a large face database efficiently. Going through demand analysis, summary design and detailed design, this thesis designed and implemented a large-scale facial-recognition surveillance system using C# and C + + programming language. The system consists of 5 subsystems, which are face data acquisition system responsible for acquiring the images through decoding video stream, comparison of face features system responsible for calculating the similarity of two facial features, load balancing system responsible for scheduling hardware resources, the Web-based database management system responsible for the adding, modifying, searching, deleting of the face library and maintaining logs of the whole system, and blacklist alarm system responsible for alarming when finding anyone suspicious. Finally, it indicates that the system corresponds with the design requirements through testing. And the system is prepared for the further practical applications.
Keyword人脸识别 视频监控 分布式 负载均衡 Face Recognition Video Surveillance Distributed System Load Balancing
Document Type学位论文
Recommended Citation
GB/T 7714
李守斌. 大规模人脸识别监控系统研究与实现[D]. 中国科学院自动化研究所. 中国科学院大学,2013.
Files in This Item:
File Name/Size DocType Version Access License
CASIA_2010E800907001(3475KB) 暂不开放CC BY-NC-SA
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[李守斌]'s Articles
Baidu academic
Similar articles in Baidu academic
[李守斌]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[李守斌]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.