期刊文献+

结合航拍图与车载图像的车辆路面位置定位 被引量:1

Positioning of Vehicle Road Position with Aerial Image and Vehicle-borne Images
下载PDF
导出
摘要 在道路交通事故司法鉴定实践中,准确定位事故车辆在路面上的位置,对事故过程还原和事故成因分析具有重要意义。引入航拍及三维激光扫描技术,运用几何映射原理,提出了一种利用航拍图像景物特征与车载视频图像景物特征相匹配的方法,对事故车辆的路面位置进行定位。该方法既无需构建一致性场景,也无需在原始现场中放置车辆;仅通过对原始现场及标定现场进行航拍,捕捉相关环境特征及车辆特征;并结合事故车辆的车载视频图像,将车辆图像放置在原始现场的航拍图中,以获取相关车辆的位置参数。结果表明,以图像中央区域特征和车前近距离特征为基础,当选择一组特征的视角约为20°时,车辆路面位置的定位效果最佳,其定位误差可控制在20 cm以内。 In the practice of forensic identification of road traffic accidents,it is of great significance to accurately locate the accident vehicle on the road and to restore accident process and analyze the cause of accident.Aerial photography and three-dimensional laser scanning technology are introduced,the principle of geometric mapping is used,a method of using aerial images to match the scene features of vehicle-borne video images is proposed to locate the road position of the accident vehicle.This method does not need to build the consistent scene,and does not need to place the vehicle in the original scene,only through the original scene and the calibration scene aerial shooting,capturing the relevant environmental characteristics and vehicle characteristics,and combined with the vehicle-borne video image of the accident vehicle to place the vehicle image in the original scene of the aerial picture,to obtain the relevant vehicle position parameters.The results show that,based on the characteristics of the central area of the image and the characteristics of the front and the front,the positioning error of the vehicle pavement position is the best when the angle of view of the selected set of features is about 20°,and the positioning errors can be controlled within 20 cm.
作者 叶长华 杨丰 刘国庆 李冬日 赵卫东 邹崇彬 YE Zhang-hua;YANG Feng;LIU Guo-qing;LI Dong-ri;ZHAO Wei-dong(School of Biomedical Engineering,Southern Medical University;School of Forensic Medicine,Southern Medical University;Guangdong Provincial Research Center of Traffic Accident Identification Engineering Technology,Guangzhou 510515,China)
出处 《科学技术与工程》 北大核心 2019年第3期239-246,共8页 Science Technology and Engineering
基金 国家自然科学基金面上项目(61771233 81772021)资助
关键词 交通工程 车辆位置 航拍图像 车载图像 司法鉴定 traffic engineering vehicle location aerial image vehicle-borne image judicial expertise
  • 相关文献

参考文献8

二级参考文献57

共引文献118

同被引文献1

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部