Vehicle Automatic Detection Algorithm from UAV Videos
Based on Shape Analysis

doi:10.3969/j.issn.1674-0696.2019.04.03

Journal of Chongqing Jiaotong University(Natural Science) ›› 2019, Vol. 38 ›› Issue (04): 15-22.DOI: 10.3969/j.issn.1674-0696.2019.04.03

• Transport+Big Data and Artificial Intelligence • Previous Articles Next Articles

Vehicle Automatic Detection Algorithm from UAV Videos Based on Shape Analysis

PENG Bo1,2,3, CAI Xiaoyu1,2, ZHOU Tao4, LI Shaobo2, ZHANG Youjie2, DUAN Lianfei5

(1.Chongqing Key Lab of Traffic System & Safety in Mountain Cities, Chongqing 400074, P. R. China; 2.College of Traffic and Transportation, Chongqing Jiaotong University, Chongqing 400074, P. R. China; 3. Key Laboratory of Urban ITS Technology Optimization and Integration Ministry of Public Security, Hefei 230088, Anhui, P. R. China; 4. Chongqing Transport Planning and Research Institute, Chongqing 401147, P. R. China; 5. Anhui Keli Information Industry Co.,Ltd, Hefei 230088, Anhui, P. R. China)

Received:2017-11-06 Revised:2018-03-05 Online:2019-04-15 Published:2019-04-15

基于形态分析的无人机视频车辆自动识别算法

彭博1,2,3，蔡晓禹1,2，周涛4，李少博2，张有节2，段连飞5

(1. 山地城市交通系统与安全重庆市重点实验室，重庆 400074； 2. 重庆交通大学交通运输学院，重庆 400074； 3. 城市交通管理集成与优化技术公安部重点实验室，安徽合肥 230088； 4. 重庆市交通规划研究院，重庆 401147； 5. 安徽科力信息产业有限责任公司，安徽合肥 230088)

作者简介:彭博(1986—)，男，四川南充人，副教授，博士，主要从事智能交通规划方面的研究。E-mail：pengbo351@126.com。通信作者：蔡晓禹(1979—)，男，四川达州人，教授，高级工程师，博士，主要从事交通规划和智能交通方面的研究。E-mail： caixiaoyu@vip.163.com。
基金资助:
国家自然科学基金青年科学基金项目（61703064）；重庆市社会事业与民生保障科技创新专项项目（cstc2015shms-ztzx30002）；重庆市教委科学研究项目（KJ1600513）；重庆市基础前沿与技术创新项目(cstc2017jcyjAX0473, cstc2018jscx-msybX0295)；城市交通管理集成优化重点实验室与山地城市交通系统安全实验室开放基金(2017KFKT01, 2018TSSMC05)

Abstract

Abstract: In order to collect continuous traffic flow information correctly and comprehensively from a regional perspective, a vehicle automatic detection method was proposed based on shape analysis aiming at UAV (unmanned aerial vehicle) videos. Firstly, a ROI (region of interest) was marked manually on the video frame, and grayscale processing was conducted on the frame. Secondly, a sub-pixel skeleton image was generated based on Canny edge detection result of ROI, and the image skeleton was decomposed and reconstructed. Then, vehicle targets were recognized through comprehensive application of morphological operations (dilation, erosion, filling and closing) and connected components shape features (area, rectangularity, major axis and minor axis of equivalent ellipse). Finally, algorithm detection and manual inspection were respectively conducted on 548 UAV video frames, and correct detection rate, repeated detection rate, missed detection rate and false detection rate for vehicle detection were calculated. Test results show that the proposed algorithm achieves higher correct detection rate (averaging 95.02%), lower repeated detection rate (averaging 2.20%), missed detection rate (averaging 2.77%) and false detection rate (averaging 8.24%). Besides, standard deviations of the correct detection rate, repeated detection rate, missed detection rate and false detection rate are 2.09%, 1.67%, 1.67% and 2.56%, respectively, which indicates that the proposed algorithm obtains smaller discrete degree of performance indexes and higher stability.

Key words: traffic engineering, vehicle detection, skeleton reconstruction, shape analysis, unmanned aerial vehicle, region of interest

摘要： 为从广域的视角准确全面地采集连续交通流信息，针对悬停无人机视频提出了基于形态分析的车辆自动识别方法。首先，人工勾画视频帧图像的感兴趣区域，并进行灰度化处理；其次，基于感兴趣区域的Canny边缘检测结果生成亚像素级骨架图像，并对图像骨架进行分解和重构处理；然后，综合应用形态学运算(膨胀、腐蚀、填充、闭运算)和连通域形态特征(面积、矩形度、等效椭圆长轴与短轴)识别车辆目标；最后，对548帧无人机视频图像分别进行算法检测和人工识别，并计算车辆识别的正检率、重检率、漏检率和错检率。结果表明：该算法具有较高的正检率(均值95.02%)，较低的重检率(均值2.20%)、漏检率(均值2.77%)和错检率(均值8.24%)；同时，正检率、重检率、漏检率和错检率标准差分别为2.09%、1.67%、1.67%和2.56%，表明算法性能指标值离散程度较小、稳定性较高。

关键词: 交通工程, 车辆检测, 骨架重构, 形态分析, 无人机, 感兴趣区域

CLC Number:

U491.1+16

PENG Bo1,2,3, CAI Xiaoyu1,2, ZHOU Tao4, LI Shaobo2, ZHANG Youjie2, DUAN Lianfei5. Vehicle Automatic Detection Algorithm from UAV Videos Based on Shape Analysis[J]. Journal of Chongqing Jiaotong University(Natural Science), 2019, 38(04): 15-22.

彭博1,2,3，蔡晓禹1,2，周涛4，李少博2，张有节2，段连飞5. 基于形态分析的无人机视频车辆自动识别算法[J]. 重庆交通大学学报（自然科学版）, 2019, 38(04): 15-22.

References

［1］ CARROLL E A, DAN B R. Using an unmanned airborne data acquisition system(ADAS) for traffic surveillance, monitoring,
and management［J］. Asme International Mechanical Engineering Congress & Exposition, 2002, 11(1): 15-26.
［2］ KIM N V, CHERVONENKIS M A. Situation control of unmanned aerial vehicles for road traffic monitoring［J］.
Modern
Applied Science, 2015, 9(5): 1-13.
［3］彭仲仁，刘晓锋，张立业，等.无人飞机在交通信息采集中的研究进展和展望［J］.交通运输工程学报，2012，12(6)：119-126.
PENG Zhongren, LIU Xiaofeng, ZHANG Liye, et al. Research progress and prospect of UAV applications in transportation
information collection［J］. Journal of Traffic and Transportation Engineering, 2012, 12(6): 119-126.
［4］ APELTAUER J, BABINEC A, HERMAN D, et al. Automatic vehicle trajectory extraction for traffic analysis from
aerial
video data［C］// International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences.
Munich,
Germany: ISPRS, 2015, XL-3/W2(3): 9-15.
［5］周雨阳，龚艺，姚琳，等.无人机广域视频的机动车交通参数计算及分析［J］.交通运输系统工程与信息，2015，15(6)：67-73.
ZHOU Yuyang, GONG Yi, YAO Lin, et al. Calculation and analysis of the traffic parameters of vehicles based on the
wide-
area drone video［J］. Journal of Transportation Systems Engineering and Information Technology, 2015, 15(6): 67-73.
［6］ KAHAKI S M. Incident and traffic-bottleneck detection algorithm in high-resolution remote sensing imagery［J］.
Journal of Information & Communication Technology, 2012, 6(2): 151-170.
［7］ KOZEMPEL K, REULKE R. Fast vehicle detection and tracking in aerial image bursts［J］. International Archives of
the
Photogrammetry Remote Sensing & Spatial Information Science, 2009, 38(3): 175-180.
［8］ CHENG Penggen, ZHOU Guoqing, ZHENG Zezhong. Detecting and counting vehicles from small low-cost UAV images
［C］//
ASPRS 2009 Annual Conference. Baltimore, Maryland, 2009.
［9］刘慧，李清泉，曾喆，等.利用低空视频检测道路车辆［J］.武汉大学学报(信息科学版)，2011，36(3)：316-320.
LIU Hui, LI Qingquan, ZENG Zhe, et al. Vehicle detection in low-altitude aircraft video［J］. Geomatics and
Information
Science of Wuhan University, 2011, 36(3): 316-320.
［10］ AZEVEDO C L, CARDOSO J L, BEN-AKIVA M, et al. Automatic vehicle trajectory extraction by aerial remote sensing
［J］
. Procedia-Social and Behavioral Sciences, 2014, 111: 849-858.
［11］张新.基于无人机视频的交通参数提取方法及应用［D］.上海：同济大学，2014.
ZHANG Xin.Method of Extracting Traffic Information in Unmanned Aerial Vehicle Videos and Its Application［D］.
Shanghai:
Tongji University, 2014.
［12］ YU Xinran, SHI Zhenwei. Vehicle detection in remote sensing imagery based on salient information and local
shape
feature［J］. Optik-International Journal for Light and Electron Optics, 2015, 126(20): 2485-2490.
［13］ RAMER U. An iterative procedure for the polygonal approximation of plane curves［J］. Computer Graphics & Image
Process, 1972, 1 (3): 244-256.
［14］王泽京，高晓阳，毕阳，等.基于等效椭圆和BP神经网络的马铃薯形状分类研究［J］.甘肃农业大学学报，2011，46(3)：131-
135.
WANG Zejing, GAO Xiaoyang, BI Yang, et al. Potato shape classification based on equivalent ellipse and BP neural
network［
J］.Journal of Gansu Agricultural University, 2011, 46(3):131-135.
［15］刘昱岗，王卓君，刘艳芳，等.基于双目视觉图像的倒车障碍物检测预处理方法［J］.重庆交通大学学报(自然科学版)，2018，
37
(3)：92-98.
LIU Yugang, WANG Zhuojun, LIU Yanfang, et al. Reversing obstacle detection based on binocular vision image［J］.
Journal
of Chongqing Jiaotong University(Natural Science), 2018, 37(3): 92-98.
［16］粟周瑜，兰全祥，袁泉，等.基于PCNN与形态学的坑洞图像边缘提取［J］.重庆交通大学学报(自然科学版)，2016，35(1)：60
-65.
SU Zhouyu, LAN Quanxiang, YUAN Quan, et al.Potholes image edge extraction based on PCNN and morphology［J］. Journal
of
Chongqing Jiaotong University(Natural Science), 2016, 35(1): 60-65.

[1]	TANG Jiqiang1,2, ZHONG Xinwei2, LIU Jian1, LI Tianrui3. Short Term Forecast of Rail Transit Passenger Flow Based on Time Series Seasonal Classification Model [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(07): 31-38.
[2]	QI Chunhua, WANG Xiaonan, ZHU Shoulin, LI Hangtian. Traffic Engineering Facility Information Threshold Based on Drivers Visual Area of Interest [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(07): 53-60.
[3]	DENG Tianmin1, ZHU Jie2, ZHU Kaijia1, QU Zhihua3. Driving Style Recognition Method Based on iForest and Bisecting K-means Combined Algorithm [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(06): 1-6.
[4]	XU Maozeng, ZHANG Li. Identification of Road Section Congestion Influence Based on Bijective Soft Decision Theory [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(06): 7-11.
[5]	WANG Tao1, XIE Sihong1, LI Wenhao1,2, LI Wenyong1. Short-Term Traffic Flow Adaptive Prediction Model Based on FFOS-ELM and PF [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(06): 21-27.
[6]	ZHANG Cheng1, WANG Chengyin1, CHEN Zhiwei2, ZHANG Zhengxiang3, ZHANG Zilei3. Mining Method of Key Nodes of Urban Road Network Based on Spatial-Flow Degree [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(06): 28-35.
[7]	WANG Yonggang, ZHANG Heng, PENG Zhipeng, ZHOU Wanqi. Analysis of Influencing Factors of Taxi Accidents Based on Structural Equation Model [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(06): 36-42.
[8]	ZHU Shunyin, LI Weiji, LIN Xiaoyuan, XIAO Wenbin, WANG Hong. Setting No Re-inspection Control Point for Tourist Vehicles in Characteristic Towns [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(06): 52-58.
[9]	PAN Binghong1,2, HU Wei1, REN Hui1, SHAN Huimin1. Truck Side-Slip Velocity Model in Curve Road Section Based on TruckSim Simulation [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(05): 38-45.
[10]	WU Wenjing, SUN Renchao, ZONG Fang, JIA Hongfei. Subjective Attitude Identification and Impact Analysis of Residents Low-Carbon Commuting Travel [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(05): 53-58.
[11]	LI Jun, DENG Yuxin. Cluster Analysis of Taxi Spatiotemporal Behavior Based on Non-negative Matrix Factorization [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(05): 59-65.
[12]	CHENG Wei1, LIU Xiang1, LEI Jianming2. Optimization of Cycle Length of Intersection Signal Based on RBF Neural Network Prediction [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(04): 13-18.
[13]	SHI Tianliang, WANG Wenguang. Information View and Particle Swarm Optimization Algorithm in Urban Traffic Congestion [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(04): 19-25.
[14]	HE Yulong1, 2, LIU Lei1,2, CHI Jiaxin1,2. Risk of Vehicle Lane Change Behavior on Highway [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(04): 26-33.
[15]	WANG Lu1, CHEN Xiaoxin2, LI Huaen1. Critical Speed of Vehicles Running Stably on Expressway in the Weather of Rain and Snow under Strong Wind [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(04): 34-40.

Vehicle Automatic Detection Algorithm from UAV Videos Based on Shape Analysis

基于形态分析的无人机视频车辆自动识别算法

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 15

Recommended Articles

Metrics