车载激光点云与序列化全景影像融合方法  

Fusion Method of Vehicle Laser Point Cloud and Serialized Panoramic Image

作　　者：卢秀山 俞家勇[2] 田茂义 刘如飞[2] 石波[2] 李鹏飞[2] Lu Xiushan;Yu Jiayong;Tian Maoyi;Liu Rufei;Shi Bo;Li Pengfei(Institute of Ocean Engineering, Shandong University of Science and Technology, Qingdao, Shandong 266590, China;College of Geomatics, Shandong University of Science and Technology, Qingdao, Shandong 266590, China)

机构地区：[1]山东科技大学海洋工程研究院,山东青岛266590 [2]山东科技大学测绘科学与工程学院,山东青岛266590

出　　处：《中国激光》2018年第5期239-246,共8页Chinese Journal of Lasers

基　　金：国家重大仪器设备开发专项(2013YQ120343);海洋公益性行业科研专项(201305034-1);测绘公益性行业科研专项(201512034)

摘　　要：利用车载移动测量系统采集的全景影像,建立全景球坐标系。将车载点云通过一系列的坐标转换,在全景球坐标系下利用球心、球面上像点和球面上物点三点共线关系,实现车载移动测量系统激光点云与全景影像的融合。全景影像拍摄盲区造成真彩点云存在"黑洞",从而导致导致路面信息缺失。针对该问题利用相邻影像重叠关系对"黑洞"进行修补。通过人工采集检查点对融合精度进行评定,结果表明,该方法融合结果正确,全景影像与激光点云融合精准,修补"黑洞"后,路面信息保留完整。We build a panoramic spherical coordinate system using the panoramic images collected by vehicle mobile measurement system.The fusion of the laser point cloud and panoramic images is realized through a series of coordinate transformation of vehicle points based on collinear principle of center of sphere,image point and object point of spherical surface.The black hole of true color point cloud caused by the blind area of panoramic images creates the road information loss,we use the overlapping relationship of adjacent images to solve the black hole.Precision evaluation is conducted via manual information collection.We use the manual collection checkpoints to evaluate the fusion accuracy,and the results show that the fusion method is correct and precision.The road information is preserved completely after repairing the black hole by accurate collection of panoramic images and laser point cloud.

关 键 词：传感器 车载移动测量系统 全景影像 球系共线方程 真彩点云 

分 类 号：P237[天文地球—摄影测量与遥感]