基于毫米波雷达和机器视觉信息融合的障碍物检测

doi:10.11959/j.issn.2096-3750.2017.00021

物联网学报 ›› 2017, Vol. 1 ›› Issue (2): 76-83.doi: 10.11959/j.issn.2096-3750.2017.00021

基于毫米波雷达和机器视觉信息融合的障碍物检测

翟光耀¹,陈蓉¹,张剑锋²,张继光³,吴澄¹,汪一鸣¹

¹ 苏州大学交通工程研究中心，江苏苏州 215000
² 苏州富欣智能交通科技有限公司，江苏苏州 215000
³ 苏州高新有轨电车有限公司，江苏苏州 215000

修回日期:2017-06-20 出版日期:2017-09-01 发布日期:2018-03-10
作者简介:翟光耀（1991-），男，苏州大学硕士生，主要研究方向为毫米波雷达、机器视觉和信息融合等。|陈蓉（1983-），女，博士，苏州大学实验师，主要研究方向为通信信号处理、时频分析理论、雷达信号处理及自适应信号处理等。|张剑锋，男，苏州富欣智能交通控制有限公司常务副总经理，苏州大学产业教授，主要研究方向为轨道交通。|张继光，男，苏州高新有轨电车有限公司高级工程师，主要研究方向为轨道交通。|吴澄（1976-），男，苏州大学副教授，主要研究方向为计算机工程、人工智能、图像处理等。|汪一鸣（1956-），女，苏州大学教授、博士生导师，中国电子学会高级会员，IEEE 会员，苏州大学通信与信息系统学科带头人之一，主要研究方向为无线通信网络、认知无线电、超宽带通信等。
基金资助:
江苏省科技厅基金资助项目(BY2015039-12);江苏省科技厅基金资助项目(KYLX_1231)

Tramway obstacles detection based on information fusion of MMV radar and machine vision

Guang-yao ZHAI¹,Rong CHEN¹,Jian-feng ZHANG²,Ji-guang ZHANG³,Cheng WU¹,Yi-ming WANG¹

¹ Transportation Engineering Research Center Soochow University,Suzhou 215000,China
² Shanghai Fuxin Intelligent Transportation Solutions Co.,Ltd.Suzhou 215000,China
³ Suzhou Gaoxin District Tramway Co.,Ltd.,Suzhou 215000,China

Revised:2017-06-20 Online:2017-09-01 Published:2018-03-10
Supported by:
Project Supported by the Science and Technology Department of Jiangsu Province(BY2015039-12);Project Supported by the Science and Technology Department of Jiangsu Province(KYLX_1231)

摘要/Abstract

摘要：

提出一种基于毫米波雷达和机器视觉传感器信息融合的障碍物检测方法。首先对毫米波雷达和摄像头进行联合标定，实现雷达与图像数据的时空同步，将雷达探测到的目标位置准确投影到图像中，进而提出一种生成雷达目标感兴趣区域的方法，同时对图像信息运用帧差法，检测图像中运动的物体，得到检测区域。最后将雷达检测区域与机器视觉检测区域进行对比，计算重合度，并根据重合度初步区分目标为行人或车辆。实验结果表明，该方法能够很好地实现毫米波雷达与机器视觉联合检测障碍物，弥补了单一传感器在障碍物检测中的不足。

关键词: 智能驾驶, 毫米波雷达, 机器视觉, 信息融合

Abstract:

A method of obstacle detection based on information fusion of millimeter wave radar and machine vision sensor was proposed.Firstly,the millimeter wave radar and camera calibration,image and radar realize spatiotemporal data synchronization,the target position was accurate projection detected by radar images,and then a method of generating radar target region of interest,and the image information using the frame difference method,image detection of moving objects,detection area was put forward.Finally,the radar detection area was compared with the machine vision detection area.The coincidence degree was calculated,and the target was divided into pedestrian or vehicle according to coincidence degree.The experimental results show that the method can detect the obstacle of millimeter wave radar and machine vision well,and make up for the shortage of single sensor in obstacle detection.

Key words: intelligent driving, millimeter wave radar, machine vision, information fusion

翟光耀,陈蓉,张剑锋,张继光,吴澄,汪一鸣. 基于毫米波雷达和机器视觉信息融合的障碍物检测[J]. 物联网学报, 2017, 1(2): 76-83.

Guang-yao ZHAI,Rong CHEN,Jian-feng ZHANG,Ji-guang ZHANG,Cheng WU,Yi-ming WANG. Tramway obstacles detection based on information fusion of MMV radar and machine vision[J]. Chinese Journal on Internet of Things, 2017, 1(2): 76-83.

图/表 16

图1

图2

图3

图4

图5

图6

图7

图8

图9

表1

图10

表2

图11

图12

图13

图14

参考文献 13

[1]	PIETZSCH S , VU T , BURLET J ,et al. Results of a precrash application based on laser scanner and short-range radars[J]. IEEE Transactions on Intelligent Transportation Systems, 2009,10(4): 584-593.
[2]	BERTOLAZZI E , BIRAL F , DA L ,et al. Supporting drivers in keeping safe speed and safe distance:the SASPENCE subproject within the European framework programme 6 integrating project Prevent[J]. IEEE Transactions on Intelligent Transportation Systems, 2010,11(3): 525-538.
[3]	郭磊, 李克强, 王建强 ,等. 一种基于特征的车辆检测方法[J]. 汽车工程, 2006,28(11): 1031-1035.
	GUO L , LI K Q , WANG J Q ,et al. A feature-based vehicle detection method[J]. Automotive Engineering, 2006,28(11): 1031-1035.
[4]	CHOI H C , PARK J M , CHOI W S ,et al. Version-based fusion of robust lane tracking and forward vehicle detection in a real driving environment[J]. International Journal of Automotive Technology, 2012,13(4): 653-669.
[5]	STOJMENOVI? M .Real time object detection in images based on an AdaBoost machine learning approach and a small training set[D]. Carleton University, 2005.
[6]	WEN C C , CHILR W C . Online boosting for vehicle detection[J]. IEEE Transactions on System,Man,and Cybernetions,Part B:Cybernetions.Published by Institute of Electrical and Electron, 2010,40(3): 892-902.
[7]	GILLNER W J , . Motion based vehicle detection motorways[C]// IEEE Intelligent Vehicle SymPosium. NewYork NY USA, 1995: 483-487.
[8]	侯德藻, 李克强, 郑四发 ,等. 汽车主动避撞系统中的报警方法及其关键技术[J]. 汽车工程, 2002,24(5): 438-444.
	HOU D Z , LI K Q , ZHENG S F ,et al. The warning algorithm and some key technologies of vehicle collision avoidance system[J]. Automotive Engineering, 2002,24(5): 438-444.
[9]	FAYAD F , CHERFAOUI V . Tracking object using a laser scanner in driving situation based on modeling target shape[C]// IEEE Intelligent Vehicles Symposium. Istanbul,Turky, 2007: 44-49.
[10]	甘志梅, 王春香, 杨明 . 基于激光雷达的车辆跟踪与识别方法[J]. 上海交通大学学报, 2009,43(6): 923-926.
	GAN Z M , WANG C X , YANG M . A method for vehicle tracking and recognition based on scanning laser radar[J]. Journal of Shanghai Jiaotong University, 2009,43(6): 923-926.
[11]	黄伟 . 基于雷达和机器视觉的车辆前方障碍物检测系统设计与实现[D]. 武汉:武汉理工大学, 2010.
	HUANG W . Design and implementation of vehicle obstacle ahead detection system based on radar and computer vision[D]. Wuhan:Wuhan University of Technology, 2010.
[12]	王亚丽 . 基于毫米波雷达与机器视觉融合的前方车辆检测研究[D]. 长春:吉林大学, 2013.
	WANG Y L . Study on decoction of ahead vehicle based on millimeter-wave radar-vision[D]. Changchun:Jilin University, 2013.
[13]	MURAT T A . Digital video processing[M]. Prentice-Hall Press, 1996.

距离/m	目标为人的重合度	目标为车的重合度
40	(30.6 ±3)%	(72.9 ±6)%
50	(29.7 ±4)%	(71.6 ±5)%
60	(30.4 ±3)%	(68.9 ±6)%
70	(28.8 ±3)%	(64.7 ±6)%
80	(27.8 ±3)%	(60.2 ±8)%
90	(26.4 ±4)%	(59.2 ±8)%
100	(25.1 ±4)%	(57.6 ±10)%

距离/m	目标为人的重合度	目标为车的重合度
40	(31.4 ±3)%	(77.6 ±8)%
50	(30.9 ±3)%	(75.8 ±8)%
60	(30.2 ±4)%	(72.9 ±7)%
70	(29.9 ±3)%	(70.0 ±6)%
80	(28.3 ±3)%	(65.3 ±8)%
90	(27.5 ±3)%	(64.7 ±7)%
100	(26.1 ±4)%	(61.7 ±9)%

基于毫米波雷达和机器视觉信息融合的障碍物检测

Tramway obstacles detection based on information fusion of MMV radar and machine vision

在线阅读

PDF下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 16

参考文献 13

相关文章 1

Metrics

推荐阅读 0