基于行车视频消失点和逆透视变换的车速测量*

doi:10.16265/j.cnki.issn 1003-3033.2021.06.017

中国安全科学学报 ›› 2021, Vol. 31 ›› Issue (6): 128-135.doi: 10.16265/j.cnki.issn 1003-3033.2021.06.017

基于行车视频消失点和逆透视变换的车速测量^*

关闯^1,2, 冯浩^**1,2 高级工程师, 张志勇^1,2 高级工程师, 姜镇飞^1,2, 张泽枫^1,2 工程师

1 司法鉴定科学研究院上海市司法鉴定专业技术服务平台,上海 200063;
2 司法部司法鉴定重点实验室,上海 200063

收稿日期:2021-03-19 修回日期:2021-05-13 出版日期:2021-06-28
作者简介:关闯 (1994—),男,河南周口人,硕士,助理工程师,主要从事于交通事故重建方面的工作。E-mail: guanc@ssfjd.cn。
基金资助:
上海市科技创新行动计划项目(20DZ2205200);中央级科研院所公益项目(GY2020Z-1,GY2020G-7)。

Vehicle speed measurement based on vanishing points of driving videos and inverse perspective mapping

GUAN Chuang^1,2, FENG Hao^1,2, ZHANG Zhiyong^1,2, JIANG Zhenfei^1,2, ZHANG Zefeng^1,2

1 Academy of Forensic Science, Shanghai Forensic Service Platform, Shanghai 200063, China;
2 Key Laboratory of Forensic Science, Ministry of Justice, Shanghai 200063, China

Received:2021-03-19 Revised:2021-05-13 Published:2021-06-28

摘要/Abstract

摘要： 为了克服行车途中相机抖动对车速测量的影响,利用行车视频消失点特性和逆透视变换特征,提出一种通用车速测量方法。首先通过检测车道线来识别获取消失点位置和车道斜率信息;其次建立相机三参数模型,并利用透视成像原理,推导出俯仰角、偏航角以及高度的计算公式;然后对逆透视变换后的正投影图像进行特征运动分析,获取自车及目标车辆速度;最后分别在自车车速为20、25、30、35、40、50 km/h以及前方目标车辆车速为10、20、30 km/h的情况下进行实车试验。结果表明:自车车速的平均误差不超过5%,前车车速的平均误差不超过7%,能够达到交通事故车速测量的要求。

关键词: 行车视频, 消失点, 逆透视变换, 车速测量, 交通事故

Abstract: In order to overcome impact of camera jitter on speed measurement during driving, a general speed measurement method was proposed based on characteristics of driving video's vanishing points and inverse perspective mapping. Firstly, positions of those points and slope information of traffic lanes were obtained by detecting and identifying lane lines. Secondly, a three parameter model of cameras was established, and calculation formulas of pitch angle, yaw angle and height were derived by using perspective imaging principle. Then, velocity of self-driving vehicle and target vehicle were obtained through analyzing feature motion of forward projection image after reverse perspective transformation. Finally, a real vehicle test was carried out under conditions of self-driving speed 20, 25, 30, 35, 40, 50 km/h and front target vehicle at 10, 20, 30 km/h respectively. The results show that the average error of self-driving speed is less than 5%, while that of front vehicle is less than 7%, which can meet requirements of traffic accident speed measurement.

Key words: driving video, vanishing point, inverse perspective mapping, speed measurement, traffic accident

中图分类号:

关闯, 冯浩, 张志勇, 姜镇飞, 张泽枫. 基于行车视频消失点和逆透视变换的车速测量^*[J]. 中国安全科学学报, 2021, 31(6): 128-135.

GUAN Chuang, FENG Hao, ZHANG Zhiyong, JIANG Zhenfei, ZHANG Zefeng. Vehicle speed measurement based on vanishing points of driving videos and inverse perspective mapping[J]. China Safety Science Journal, 2021, 31(6): 128-135.

参考文献

[1] 徐丹, 代勇, 纪军红. 基于卷积神经网络的驾驶人行为识别方法研究[J]. 中国安全科学学报, 2019, 29 (10):12-17.
XU Dan, DAI Yong, JI Junhong. Research on driver behavior recognition method based on convolutional neural network[J]. China Safety Science Journal, 2019, 29 (10):12-17.
[2] 刘通,付锐,张士伟,等.车辆典型危险行驶状态识别与检测研究进展[J].中国安全科学学报,2017,27(10):32-37.
LIU Tong, FU Rui, ZHANG Shiwei, et al. Progress in research on identification and detection of vehicle typical hazardous driving states[J]. China Safety Science Journal, 2017,27(10):32-37.
[3] HAN J, HEO O, PARK M, et al. Vehicle distance estimation using a mono-camera for FCW/AEB systems[J]. International Journal of Automotive Technology, 2016, 17(3):483-491.
[4] ARENADO M I, TORRE-FERRERO C ,RENTERIA L A. et al. Monovision based vehicle detection, distance and relative speed measurement in urban traffic[J]. IET Intelligent Transport Systems,2014, 8(8):655-664.
[5] 余厚云, 张为公. 基于单目视觉的跟驰车辆车距测量方法[J]. 东南大学学报:自然科学版, 2012, 42(3):154-158.
YU Houyun, ZHANG Weigong. Method of vehicle distance measurement for following car based on monocular vision[J]. Journal of Southeast University: Natural Science Edition, 2012, 42(3):542-546.
[6] 陈荣保, 赵聃, 王乾隆. 基于图像处理的前方行驶车辆速度测量方法[J]. 传感器与微系统, 2018, 37(4): 17-19.
CHEN Rongbao, ZHAO Dan, WANG Qianlong. Measurement method of front vehicle speed based on image processing[J]. Transducer and Microsystem Technologies, 2018, 37(4): 17-19.
[7] KANEKO H, MORIMOTO M, FUJII K. Vehicle speed estimation by in vehicle camera[C].World Automation Congress (WAC), 2012. IEEE, 2012:1-6.
[8] 刘军, 后士浩, 张凯,等. 基于单目视觉车辆姿态角估计和逆透视变换的车距测量[J]. 农业工程学报, 2018, 34(13):70-76.
LIU Jun, HOU Shihao, ZHANG Kai, et al. Vehicle distance measurement with implementation of vehicle attitude angle estimation and inverse perspective mapping based on monocular vision[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2018, 34(13): 70-76.
[9] LIN Chienchuan, WANG Mingshi. Vehicle speeding early warning model using frame feature detection and HMM[C]. IEEE International Conference on Consumer Electronics,2011:DOI:10.1109/ISCE.2011.5973824.
[10] 周劲草,魏朗,乔洁.面向雾天的双先验兴趣区域车道线识别算法研究[J].中国安全科学学报,2016,26(8):128-133.
ZHOU Jincao, WEI Lang, QIAO Jie. Study on fog-oriented lane detection method based on dual-heuristic region of interest[J]. China Safety Science Journal, 2016,26(8):128-133.
[11] 陈政宏, 李爱娟, 王希波, 等. 基于改进Hough变换的结构化道路车道线识别[J]. 科学技术与工程, 2020, 20(26): 10 829-10 834.
CHEN Zhenghong, LI Aijuan, WANG Xibo, et al. Structured road lane line recognition based on improved Hough transformation[J]. Science Technology and Engineering, 2020, 20(26) : 10 829-10 834.
[12] 李亚娣, 黄海波, 李相鹏,等. 基于Canny算子和Hough变换的夜间车道线检测[J]. 科学技术与工程, 2016, 16(31): 234-237.
LI Yadi, HUANG Haibo, LI Xiangpeng, et al. Nighttime lane markings detection based on Canny operator and Hough transform[J]. Science Technology and Engineering, 2016, 16(31): 234-237.
[13] ZHANG Daiming , FANG Bin , YANG Weibin , et al. Robust inverse perspective mapping based on vanishing point[C].International Conference on Security. IEEE, 2014: 458-463.
[14] 王永忠, 文成林. 基于Haar纹理的非结构化道路消失点检测[J]. 中国图象图形学报, 2013, 18(4):382-391.
WANG Yongzhong, WEN Chenglin. Vanishing point detection of unstructured road based on Haar texture[J]. Journal of Image and Graphics, 2013, 18(4):382-391.
[15] PHUNG S L , LE M C , BOUZERDOUM A . Pedestrian lane detection in unstructured scenes for assistive navigation[J]. Computer Vision & Image Understanding, 2016, 149:186-196.
[16] LIN Chinteng , SHEN Tzukuei , SHOU Yuwen . Construction of fisheye lens inverse perspective mapping model and its applications of obstacle detection[J]. Journal on Advances in Signal Processing, 2010, 2010(1): 1-23.

基于行车视频消失点和逆透视变换的车速测量^*

Vehicle speed measurement based on vanishing points of driving videos and inverse perspective mapping

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	胡立伟, 贺雨, 侯智, 张瑞杰, 陈琛, 刘冰. 山区高速公路交通事故风险多维度耦合研究[J]. 中国安全科学学报, 2024, 34(5): 17-27.
[2]	王卓伦, 林岩. 中国网约车服务对道路交通事故的影响研究[J]. 中国安全科学学报, 2024, 34(2): 209-216.
[3]	陈娇娜, 靳引利, 陶伟俊, 李道峰. 基于文本信息的高速公路事故持续时间中介效应研究[J]. 中国安全科学学报, 2023, 33(4): 155-162.
[4]	付姗姗, 张悦, 席永涛, 翁金贤. 多因素耦合下长江口水域交通事故致因链分析[J]. 中国安全科学学报, 2023, 33(3): 60-67.
[5]	王亮. 基于贝叶斯网络的铁路交通事故预警研究[J]. 中国安全科学学报, 2022, 32(S1): 134-139.
[6]	熊晓夏, 刘擎超, 沈钰杰, 蔡英凤, 陈龙. 基于LSTM-BF的高速公路交通事故风险模型[J]. 中国安全科学学报, 2022, 32(5): 170-176.
[7]	杨利超, 刘敬贤, 刘钊, 刘轶伟. 价值衰减下水上交通事故致因关联规则挖掘[J]. 中国安全科学学报, 2022, 32(10): 127-134.
[8]	戢晓峰, 乔新, 普永明, 卢梦媛, 郝京京. 地铁站周边建成环境对交通事故风险的影响机制[J]. 中国安全科学学报, 2022, 32(10): 162-170.
[9]	黄常海, 沈佳, 朱冉超, 齐绪存, 郑菲, 陆浩. 基于C5.0决策树的船舶交通事故致因分析模型及应用[J]. 中国安全科学学报, 2022, 32(10): 90-99.
[10]	韩天园, 田顺, 吕凯光, 李旋, 张佳涛, 魏朗. 基于文本挖掘的重特大交通事故成因网络分析^*[J]. 中国安全科学学报, 2021, 31(9): 150-156.
[11]	刘文龙. 由一起诉讼案件谈对《铁路交通事故认定书》的认识[J]. 中国安全科学学报, 2020, 30(S1): 43-47.
[12]	吴佩洁, 孟祥海, 曹梦迪. 城市交通事故多发点鉴别与时空模式挖掘[J]. 中国安全科学学报, 2020, 30(11): 127-133.
[13]	李末, 燕飞, 牛儒, 宋晗炜, 王艳梅, 向楠楠. 基于安全指标监控的事故致因分析方法^*[J]. 中国安全科学学报, 2019, 29(S2): 35-40.
[14]	陆欢, 戢晓峰, 杨文臣, 胡澄宇. 高原山区公路环境下交通事故形态致因分析[J]. 中国安全科学学报, 2019, 29(5): 44-49.
[15]	胡立伟, 何越人, 佘天毅, 孟玲, 杨锦青. 高速公路应急救援中心选址优化模型[J]. 中国安全科学学报, 2019, 29(5): 145-150.