[1]陆 飞,沈世斌,苏晓云,等.基于改进Mask R-CNN的交通监控视频车辆检测算法[J].南京师范大学学报(工程技术版),2020,(04):044-50.[doi:10.3969/j.issn.1672-1292.2020.04.007]
 Lu Fei,Shen Shibin,Su Xiaoyun,et al.Vehicle Detection Algorithm Based on Improved Mask R-CNNin Traffic Surveillance Video[J].Journal of Nanjing Normal University(Engineering and Technology),2020,(04):044-50.[doi:10.3969/j.issn.1672-1292.2020.04.007]
点击复制

基于改进Mask R-CNN的交通监控视频车辆检测算法
分享到:

南京师范大学学报(工程技术版)[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2020年04期
页码:
044-50
栏目:
计算机科学与技术
出版日期:
2020-12-15

文章信息/Info

Title:
Vehicle Detection Algorithm Based on Improved Mask R-CNNin Traffic Surveillance Video
文章编号:
1672-1292(2020)04-0044-07
作者:
陆 飞12沈世斌13苏晓云12谢 非123章 悦1刘益剑123
(1.南京师范大学电气与自动化工程学院,江苏 南京 210023)(2.江苏省三维打印装备与制造重点实验室,江苏 南京 210023)(3.南京智能高端装备产业研究院有限公司,江苏 南京 210023)
Author(s):
Lu Fei12Shen Shibin13Su Xiaoyun12Xie Fei123Zhang Yue1Liu Yijian123
(1.School of Electrical and Automation Engineering,Nanjing Normal University,Nanjing 210023,China)(2.Jiangsu Key Laboratory of 3D Printing Equipment and Manufacturing,Nanjing 210023,China)(3.Nanjing Industry Institute for Advanced Intelligent Equipment Co.,Ltd.,Nanjing 210023,China)
关键词:
目标检测交通监控Mask R-CNN掩码预测
Keywords:
target detectiontraffic surveillanceMask R-CNNmask prediction
分类号:
TP391
DOI:
10.3969/j.issn.1672-1292.2020.04.007
文献标志码:
A
摘要:
针对交通监控视频车辆检测常易受到遮挡导致目标车辆出现漏检或误检的问题,提出一种基于改进Mask R-CNN的交通监控视频车辆检测算法. 采用基于bottleneck结构的主干网络,提高主干网络提取特征的能力; 通过基于预测mask分数的掩码分支,融合目标的类别分数和掩码质量分数,提高车辆的掩码质量; 通过基于Arcface Loss的目标检测损失函数设计,提高不同特征之间的可判别性,提高目标的检测精度. 实验结果表明,改进的Mask R-CNN模型可更好地检测到被遮挡的车辆,目标车辆的检测精度超过Faster R-CNN、YOLO v3和Mask R-CNN模型,可解决目标车辆漏检或误检问题.
Abstract:
Aiming at the problem of missing detection or wrong detection of target vehicles caused by occlusion in traffic surveillance video,an improved vehicle detection algorithm based on Mask R-CNN traffic surveillance video is proposed. Firstly,the backbone network based on the bottleneck structure is used to improve the ability of extracting features from the backbone network. Then,the mask branch based on the predicted mask score is used to fuse the target’s category score and mask quality score to improve the vehicle’s mask quality. Finally,the target detection loss function based on Arcface Loss can improve the discriminability between different features and improve the detection accuracy of the target. The experimental results show that the improved Mask R-CNN model can better detect the shielded vehicle,and that the detection accuracy of the target vehicle is higher than those of the Faster R-CNN,YOLO v3 and Mask R-CNN model,thus solving the problem of missing or wrong detection of the target vehicle.

参考文献/References:

[1] 沈连丰,张瑞,朱亚萍,等. 面向自动驾驶的车辆精确实时定位算法[J]. 电子与信息学报,2020,42(1):28-35.
[2]ANALA M R,MAKKER M,ASHOK A. Anomaly detection in surveillance videos[C]//2019 26th International Conference on High Performance Computing,Data and Analytics Workshop. Hyderabad,India:IEEE Computer Society,2019:93-98.
[3]GIRSHICK R,DONAHUE J,DARRELL T,et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//The IEEE Conference on Computer Vision and Pattern Recognition(CVPR). Columbus,USA:IEEE Computer Society,2014:580-587.
[4]UIJLINGS J R R,SANDE K E A,GEVERS T,et al. Selective search for object recognition[J]. International Journal of Computer Vision,2013,104(2):154-171.
[5]HE K,YU X. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence,2015,37(9):1904-1916.
[6]GIRSHICK R B. Fast R-CNN[C]//The IEEE International Conference on Computer Vision(ICCV). Santiago,Chile:IEEE Computer Society,2015:1440-1448.
[7]REN S,HE K,GIRSHICK R,et al. Faster R-CNN:towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence,2015,39(6):1137-1149.
[8]ZHANG X,LI B,HU H. Scale-aware hierarchical loss:a multipath RPN for multi-scale pedestrian detection[C]//IEEE Visual Communications and Image Processing(VCIP). Petersburg,USA:IEEE Computer Society,2017:1-4.
[9]REDMON J,DIVVALA S K,GIRSHICK R B,et al. You only look once:Unified,real-time object detection[C]//IEEE International Conference on Computer Vision and Pattern Recognition. Las Vegas,USA:IEEE Computer Society,2016:779-788.
[10]HE K,GKIOXARI G,DOLLAR P,et al. Mask R-CNN[C]//IEEE International Conference on Computer Vision and Pattern Recognition. Hawaii,USA:IEEE computer Society,2017:2961-2969.
[11]LI Y,QI H,DAI J. Fully convolutional instance-aware semantic segmentation[C]//IEEE Conference on Computer Vision and Pattern Recognition. Hawaii,USA:IEEE Computer Society,2017:4438-4446.
[12]陈泽,叶学义,钱丁炜,等. 基于改进的Faster R-CNN小尺度行人检测[J]. 计算机工程,2020,46(9):226-232.
[13]江昆鹏,闫洪涛,杨红卫,等. 改进Mask R-CNN的细粒度车型识别算法[J]. 软件,2020,41(3):1-5.
[14]朱有产,王雯瑶. 基于改进Mask R-CNN的绝缘子目标识别方法[J]. 微电子学与计算机,2020,37(2):69-74.
[15]石杰,周亚丽,张奇志. 基于改进Mask R-CNN和Kinect的服务机器人物品识别系统[J]. 仪器仪表学报,2019,40(4):216-228.
[16]马素刚,赵祥模,侯志强,等. 一种基于ResNet网络特征的视觉目标跟踪算法[J]. 北京邮电大学学报,2020,43(2):129-134.
[17]LIU W,WEN Y,YU Z,et al. SphereFace:deep hypersphere embedding for face recognition[C]//IEEE Conference on Computer Vision and Pattern Recognition(CVPR). Hawaii,USA:IEEE Computer Society,2017:6738-6746.
[18]DENG J,GUO J,ZAFEIRIOU S. ArcFace:additive angular margin loss for deep face recognition[C]//IEEE Conference on Computer Vision and Pattern Recognition(CVPR). Los Angeles,USA:IEEE Computer Society,2019:4685-4694.

相似文献/References:

[1]程显毅,胡海涛,季国华,等.基于深度学习监控场景下的多尺度目标检测算法研究[J].南京师范大学学报(工程技术版),2018,18(03):033.[doi:10.3969/j.issn.1672-1292.2018.03.005]
 Cheng Xianyi,Hu Haitao,Ji Guohua,et al.Research on Algorithm of Multi-Scale Target DetectionBased on Deep Learning in Monitoring Scenario[J].Journal of Nanjing Normal University(Engineering and Technology),2018,18(04):033.[doi:10.3969/j.issn.1672-1292.2018.03.005]

备注/Memo

备注/Memo:
收稿日期:2020-06-11.
基金项目:国家自然科学基金项目(61601228、41974033、61803208)、江苏省自然科学基金项目(BK20161021、BK20180726)、江苏省高校自然科学基金项目(17KJB510031).
通讯作者:沈世斌,高级实验员,研究方向:嵌入式系统、目标检测与跟踪、机器视觉与图像处理. E-mail:63018@njnu.edu.cn
更新日期/Last Update: 2020-12-15