一种分布式人脸识别方法及性能优化

闵卫东; 石杰; 韩清; 王玮

doi:10.3788/OPE.20172503.0779

您当前的位置：

首页 >

文章列表页 >

一种分布式人脸识别方法及性能优化

信息科学 | 更新时间：2020-07-07

- 一种分布式人脸识别方法及性能优化
- A distributed face recognition method and performance optimization
- 光学精密工程 2017年25卷第3期页码：779-785
- 作者机构：
  
  1.天津工业大学计算机科学与软件学院, 天津 300387
  2.南昌大学信息工程学院, 江西南昌 330031
- 作者简介：
  
  闵卫东 (1966－), 男, 江西会昌人, 博士, 教授, 博士生导师, 1989年、1991年、1995年于清华大学分别获得学士、硕士和博士学位, 1995年至2014年在加拿大Alberta大学和Corel等跨国公司从事研究工作, 主要从事图形图像处理、图形学、分布式系统、智慧城市等的研究。E-mail:minweidong@ncu.edu.cn MIN Wei-dong, E-mail:minweidong@ncu.edu.cn
  [ "石杰 (1990－), 女, 内蒙古赤峰人, 硕士研究生, 2013年于天津工业大学获得学士学位, 主要研究分布式人脸识别及其性能优化, 分布式网络管理。E-mail:714491893@qq.com" ]
- 基金信息：
  
  国家自然科学基金资助项目(61302127);江西省自然科学基金重大项目(20161ACB20004)
- DOI：10.3788/OPE.20172503.0779
  中图分类号： TP394.1
- 收稿日期：2016-12-21，
  
  录用日期：2017-1-15，
  
  纸质出版日期：2017-03-25
- 稿件说明：
移动端阅览
闵卫东, 石杰, 韩清, 等. 一种分布式人脸识别方法及性能优化[J]. 光学精密工程, 2017,25(3):779-785.

Wei-dong MIN, Jie SHI, Qing HAN, et al. A distributed face recognition method and performance optimization[J]. Optics and precision engineering, 2017, 25(3): 779-785.
闵卫东, 石杰, 韩清, 等. 一种分布式人脸识别方法及性能优化[J]. 光学精密工程, 2017,25(3):779-785. DOI： 10.3788/OPE.20172503.0779.

Wei-dong MIN, Jie SHI, Qing HAN, et al. A distributed face recognition method and performance optimization[J]. Optics and precision engineering, 2017, 25(3): 779-785. DOI： 10.3788/OPE.20172503.0779.

摘要

传统的集中式人脸识别方法在时间效率和可扩展性等方面存在不足，已经不能满足大规模实时人脸识别的需求。针对这个技术瓶颈，本文提出了一种分布式人脸识别方法，该模型由多个代理和一个服务器组成。代理能够同时对多个视频中的行人进行检测、跟踪以及特征提取，服务器则对视频中的行人执行识别操作。针对代理处理的任务分布不均而导致处理视频时间过长、任务量过大引起的CPU利用率激增问题，设计了代理的负载均衡来进行性能优化。利用代理统计处理的视频总数及每个视频中的行人数，并将统计数据发送给服务器。服务器通过负载均衡将视频重新分配给每个代理进行人脸识别。实验结果证明，分布式人脸识别有效地提高了人脸识别方法的效率和可扩展性。对一些较为极端的实验例子，经过性能优化后，代理中最大的CPU利用率降低近40%，有效地缓解了时间延迟问题。

Abstract

Since the scale of video to be monitored and processed by face recognition system has been increasing

the traditional centralized face recognition methods are insufficient in time efficiency and scalability and no longer able to meet the needs of large scale and real-time face recognition. Aimed at such technical bottleneck

a distributed face recognition method was proposed. The model consisted of several agents and one server

in which the agent was able to detect

trace and extract features of pedestrians in several videos at the same time

while the server was able to identify those pedestrians in videos. Subject to the problems caused by uneven distribution of agent processing tasks

including long processing time

too much tasks and CPU usage explosion

a load balance of agent was designed for performance optimization. Firstly

the agent was used to count the total videos to be processed and number of pedestrians in each video

then

statistical data was sent to the server through the agent and at last

the server will re-allocate the videos to each agent for face recognition by the load balance. The results show that the distributed face recognition method can effectively promote the efficiency and scalability of face recognition methods. For those extreme cases

after performance optimization

the maximum CPU using rate in agent has been declined by approximate 40%

which could effectively alleviate the time delay.

关键词

Keywords

references

WU Y W, AI X Y. An improvement of face detection using AdaBoost with color information [C]. International Colloquium on Computing, Communication, Control, and Management. IEEE Computer Society, Washington, DC, USA: ISECS, 2008:317-321.

BELAROUSSI R, MILGRAM M. A comparative study on face detection and tracking algorithms [J]. Expert Systems with Applications, 2012, 39(8):7158-7164.

佘九华, 王敬东, 李鹏.基于Camshift的人脸跟踪算法[J].计算机技术与发展, 2008, 18(9):12-15.

SHE J H, WANG J D, LI P. Face tracking algorithm based on Camshift [J]. Computer Technology and Development , 2008, 18(9):12-15.(in Chinese)

VILAPLANA V, MARQUES F. Region-based mean shift tracking: application to face tracking [C]. IEEE International Conference on Image Processing, Barcelona, Spain: ICIP, 2008:2712-2715.

NAIR B M, FOYTIK J, TOMPKINS R, et al .. Multi-pose face recognition and tracking system [J]. Complex Adaptive Systems , 2011, 6(1):381-386.

FOYTIK J, SANKARAN P, ASARI V. Tracking and recognizing multiple faces using kalman filter and modular PCA [J]. Procedia Computer Science , 2011, 6(1):256-261.

JIANG C H, SU G D, LIU X D. A distributed parallel system for face recognition [C]. International Conference on Parallel and Distributed Computing, Applications and Technologies, Chengdu, China: PDCAT, 2003:797-800.

ZHANG Z, GUO Y, SONG G. A distributed face recognition framework based on data fusion [J]. International Journal of Database Theory & Application , 2014, 7(4): 87-98.

YAN Y J, OSADCIW L A. Distributed wireless face recognition system [C] . Electronic Imaging, International Society for Optics and Photonics, San Jose, America: SPIE, 2008:68200A-68200A-12.

RAZZAK M I, KHAN M K, ALGATHBAR K, et al .. Energy efficient distributed face recognition in wireless sensor network [J]. Wireless Personal Communications , 2011, 60(3):571-582.

RAJESHWARI J, KARIBASAOOA K.Face recognition in video systems on homogeneous distributed systems [J]. International Journal of Advanced Computer and Mathematical Sciences , 2013, 4(1): 143-147.

WANG L, CAI J C, LI M. An adaptive multi-population genetic algorithm for job-shop scheduling problem [J]. Advances in Manufacturing, 2016, 4(2):142-149.

LU Q, LI S, ZHANG W, et al .. A genetic algorithm-based job scheduling model for big data analytics [J]. Eurasip Journal on Wireless Communications & Networking , 2016, 2016(1):1-9.

NOURZADEH R, EFFATPARVAR M. A genetic-fuzzy algorithm for load balancing in multiprocessor systems [J]. International Journal of Computer Applications , 2014, 101(10):39-42.

TED S, KENNETH N B.Wireless LAN load-balancing with genetic algorithms [J]. Knowledge-Based Systems , 2009, 22(7):529-534.

JIAN A, CHAUHARI N S.Genetic algorithm based concept design to optimize network load balance [J]. Ictact Journal on Soft Computing , 2012, 2(4):357-360.

DASGUPTA K, MANDAL B, DUTTA P, et al .. A Genetic Algorithm (GA) based load balancing strategy for cloud computing [J]. Procedia Technology , 2013, 10(2):340-347.

MIN W D, LIU Y H, KE Y Z, et al .. Using particle swarm optimization algorithm to improve multi-agents network management [J]. Journal of Computational Information Systems , 2014, 10(2):739-746.

劳达宝, 周维虎, 李万红, 等.基于遗传算法的柱面光栅测角技术研究[J].红外与激光工程, 2015, 44(7):2182-2188.

LAO D B, ZHOU W H, LI W H.Cylindrical grating angle measurement technology based on genetic algorithm [J]. Infrared and Laser Engineering , 2015, 44(7):2182-2188.(in Chinese)

郭天太, 洪博, 潘增荣, 等.基于改进的SVM在矿井气体定量分析中的应用[J].红外与激光工程, 2016, 45(6):203-210.

GUO T T, HONG B, PAN Z R, et al .. Application of improved SVM in quantitative analysis of mine gas concentration [J]. Infrared and Laser Engineering , 2016, 45(6):203-210.(in Chinese)

浏览量

369

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

单光路拉曼分布式光纤测温系统

融合卷积块注意力模块和Siamese神经网络的人脸识别算法

基于内容特征提取的短波红外-可见光人脸识别

联合Gabor降维特征与奇异值特征的人脸识别

三维到二维:人脸本征形状描述图