|本期目录/Table of Contents|

[1]晏 栋,孙作雷,张 仓,等.基于ACF算法的行人检测研究[J].电子设计工程,2017,(17):172-175.
 YAN Dong,SUN Zuo-lei,ZHANG Cang,et al.Pedestrian detection based on ACF algorithm[J].SAMSON,2017,(17):172-175.
点击复制

基于ACF算法的行人检测研究(PDF)
分享到:

《电子设计工程》[ISSN:1674-6236/CN:61-1477/TN]

卷:
期数:
2017年17期
页码:
172-175
栏目:
网络与通信
出版日期:
2017-09-05

文章信息/Info

Title:
Pedestrian detection based on ACF algorithm
文章编号:
1674-6236(2017)17-0172-04
作者:
晏 栋孙作雷张 仓黄冠明
(上海海事大学 上海 201306)
Author(s):
YAN DongSUN Zuo-leiZHANG CangHUANG Guan-ming
(Shanghai Maritime University,Shanghai 201306,China)
关键词:
行人检测 关联特征 边缘特征 突出特征 分类器
Keywords:
pedestrian detection correlation feature edgebox feature salient feature
分类号:
TN99
DOI:
-
文献标志码:
A
摘要:
近年来,关于行人检测领域的研究,人们提出了各种各样的方法。其中,多样化的特征和高效的分类器是这些研究的关键,本文就是在这两个方面找到一种更高效的方法来提高性能。我们的行人检测方法是将3种特征关联在一起,包括负特征(objectness features)、突出特征(salient feature)和边缘特征(edgebox feature)。与此同时,我们改进分类器的架构以提升性能。通过Caltech-USA 和 INRIA这两个数据库对模型进行训练,相比于一般的检测模型,我们的准确率提高了20%。
Abstract:
In recent years, various methods in the field of pedestrian detection have been proposed.Among them, a variety of features and efficient classifier is the key to these studies, this paper is to find a more efficient way in these two areas to improve performance.Our pedestrian detection method is associated with three features, including the objectness features, salient feature and edgebox feature . At the same time, we have improved the classification architecture to improve performance.Caltech-USA and INRIA two databases train the model, compared to the general detection model, our accuracy rate increased by 20%.

参考文献/References:

[1] Benenson, Rodrigo, et al. "Pedestrian detection at 100 frames per second." Computer Vision and Pattern Recognition (CVPR)[C]// 2012 IEEE Conference on. IEEE, 2012.[2] Sermanet, Pierre, et al.Pedestrian detection with unsupervised multi-stage feature learning[C].Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2013.[3] Ouyang, Wanli,Xiaogang Wang.Joint deep learning for pedestrian detection[C]// Proceedings of the IEEE International Conference on Computer Vision. 2013.[4] Benenson, Rodrigo.Ten years of pedestrian detection, what have we learned[C]// European Conference on Computer Vision. Springer International Publishing, 2014.[5] Dollár, Piotr, Ron Appel, and Wolf Kienzle.Crosstalk cascades for frame-rate pedestrian detection[J].Computer Vision-ECCV 2012. Springer Berlin Heidelberg, 2012:645-659.[6] Dollar, Piotr. Pedestrian detection: An evaluation of the state of the art[J].IEEE transactions on pattern analysis and machine intelligence 34(4),2012:743-761.[4] Benenson, Rodrigo.Ten years of pedestrian detec-tion, what have we learned[C]// European Confe-rence on Computer Vision. Springer International Publishing,2014.[7] 李娟,邵春福,杨励雅.基于混合高斯模型的行人检测方法[J].吉林大学学报: 工学版, 2011,41(1): 41-45.[8] 种衍文,匡湖林,李清泉.一种基于多特征和机器学习的分级行人检测方法[J].自动化学报, 2012,38(3):375-381.[9] 姚雪琴,李晓华,周激流.基于边缘对称性和 HOG 的行人检测方法[J].计算机工程, 2012,38(5): 179-182.[10]孙锐,陈军,高隽.基于显著性检测与 HOG-NMF 特征的快速行人检测方法[J].电子与信息学报, 2013,35(8):1921-1926.[11]付洋,宋焕生,陈艳, 等.一种基于视频的道路行人检测方法[J].电视技术,2012,36(13):140-144.[12]YAN, Jun-jie.Robust multi-resolution pedestrian detection in traffic scenes[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2013.[13]Ouyang, Wanli,Xiaogang Wang.Single-pedestrian detection aided by multi-pedestrian detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2013.[14]Luo, Ping.Switchable deep network for pedestrian detection[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2014.[15]Marin, Javier.Random forests of local experts for pedestrian detection[C]// Proceedings of the IEEE International Conference on Computer Vision,2013.[16]Zhang Shan-shan, Christian Bauckhage,Armin B. Cremers.Informed haar-like features improve pedestrian detection[C]// Proceedings of the IEEE conference on computer vision and pattern recognition,2014.

相似文献/References:

[1]吴 陈,杨镕华.基于垂直数据格式频繁闭项集的选择性集成算法的研究[J].电子设计工程,2016,(19):69.
 WU Chen,YANG Rong-hua.Research of selective ensemble besed on vertical data and closed pattern[J].SAMSON,2016,(17):69.
[2]赵 晨,孟朝晖.基于彩色和红外特征的多模式行人检测方法[J].电子设计工程,2016,(07):55.
 ZHAO Chen,MENG Zhao?hui.Multi?鄄pattern methods to pedestrian detection based on the color and infrared features[J].SAMSON,2016,(17):55.
[3]胡庆新,王 磊.基于多特征的红外图像行人检测[J].电子设计工程,2016,(04):182.
 HU Qing-xin,WANG Lei.Pedestrian detection in infrared images based on multi-features[J].SAMSON,2016,(17):182.
[4]张星龙,冯 全.采用透视变换提高车辆识别准确率的方法研究[J].电子设计工程,2015,(01):191.
 ZHANG Xing-long,FENG Quan.A method to improve the accuracy of vehicle identification by perspective transformation[J].SAMSON,2015,(17):191.
[5]张艳丽.基于IKSVM的多特征行人检测算法[J].电子设计工程,2014,(23):177.
 ZHANG Yan-li.Multiple features pedestrian detection algorithm based IKSVM[J].SAMSON,2014,(17):177.
[6]熊 聪,王文武.基于DPM模型的行人检测技术的研究[J].电子设计工程,2014,(23):172.
 XIONG Cong,WANG Wen-wu.Research on pedestrian detection based on DPM[J].SAMSON,2014,(17):172.
[7]蔡苏亚,刘 璟.基于计算机仿真的行人检查算法研究[J].电子设计工程,2014,(23):51.
 CAI Su-ya,LIU Jing.Pedestrian detection algorithm research based on computer simulation[J].SAMSON,2014,(17):51.
[8]安恒煊,张学习,李 超,等.一种基于AdaBoost人脸检测算法在Android平台的实现[J].电子设计工程,2014,(08):126.
 AN Heng-xuan,ZHANG Xue-xi,LI Chao,et al.A face detection based on AdaBoost algorithm realize on the Android platform[J].SAMSON,2014,(17):126.
[9]徐 好,陈淑荣. 基于HOG和肤色特征的快速行人检测[J].电子设计工程,2013,(12):108.
 XU Hao,CHEN Shu-rong. Based on HOG and skin color characteristics of rapid pedestrian detection[J].SAMSON,2013,(17):108.
[10]何保荣,李建荣,王超智. 基于分类算法的校园网络入侵检测系统研究[J].电子设计工程,2013,(12):12.
 HE Bao-rong,LI Jian-rong,WANG Chao-zhi. Study of the intrusion detection system in campus network based on the classification algorithm[J].SAMSON,2013,(17):12.

备注/Memo

备注/Memo:
收稿日期:2016-08-08 稿件编号:201608063作者简介:晏 栋(1991—),男,湖北随州人,硕士。研究方向:行人检测。
更新日期/Last Update: 2017-09-07