摘要
针对单一利用点云数据修复空洞而缺乏真实性的问题,提出了一种结合全景影像的车载点云空洞修复方法。该方法首先探讨了基于全景球共线方程的车载点云与全景影像精配准,然后构建了三角格网提取点云空洞范围,针对全景影像变形问题,采用了球面透视模型进行纠正,最后研究了基于SIFT算法的区域种子点密集增长算法,利用前方交会法生成填补点云。实验表明,该方法生成的填补点云具有真实性和完整性,将填补点云与原始点云融合,实现了车载点云空洞修复。
Aiming at the problem of the lack of authenticity by using point cloud data to repair the holes alone,a vehiclemounted point cloud hole repair method combined with panoramic images is proposed.This method firstly discusses the precise registration of the vehicle-mounted point cloud and the panoramic image based on the collinear equation of the panoramic sphere,and then constructs a triangular grid to extract the point cloud holes range.Aiming at the problem of panoramic image deformation,a spherical perspective model is used to correct it.Finally,the area seed point dense growth algorithm based on the SIFT algorithm is studied,and the forward intersection method is used to generate the filling point cloud.Experiments show that the filling point cloud generated by this method is authentic and complete,and the filling point cloud is merged with the original point cloud,the vehicle-mounted point cloud hole repair is realized.
作者
芮明
姚卫华
马兵权
刘东海
RUI Ming;YAO Wei-hua;MA Bing-quan;LIU Dong-hai(Changzhou Surveying and Mapping Institute,Changzhou Jiangsu 213002,China;Changzhou Geographic Information Intelligent Technology Center,Changzhou Jiangsu 213002,China)
出处
《现代测绘》
2023年第2期1-5,共5页
Modern Surveying and Mapping
基金
江苏省测绘地理信息科研项目(JSCHKY201818)。
关键词
车载点云
全景影像
点云配准
变形纠正
空洞填补
vehicle-mounted point
panoramic images
point cloud registration
deformation correction
hole filling
