摘要
针对城市复杂立交桥的空间特性,提出集成激光雷达数据和遥感影像实施立交桥的快速自动检测的技术流程和关键技术。首先提出利用区域增长方法改进的渐进加密三角网,有效剔除地面点云,避免对立交检测的干扰;然后采用模型驱动的分割算法,检测立交桥点云;采用基于自适应扫描带算法检测立交桥中心线,克服点云采样密度变化,确保提取过程和结果稳健;之后利用遥感影像边缘精化立交桥轮廓;最后利用模型平均曲率检测修复断裂立交桥单体。利用实际数据验证了整个流程和关键技术的可行性和有效性。
An automatic method for overpass detection with LiDAR data and remote sensing image is presented, according to spatial characters of urban complex overpass. First of all, the ground point cloud is filtered effectively with advanced adaptive TIN filter; secondly, viaduct point cloud is segmented with overpass model based strategy; thirdly, the center line of the overpass is detected with adaptive scan gap algorithm; then the boundary of overpass is refined with fusion of high resolution remote sensing image; at last, the occlusions in overpass are detected and repaired based on mean curvature of overpass model. The feasibility and performance of the suggested techniques are verified through experimental results with real data.
出处
《测绘学报》
EI
CSCD
北大核心
2012年第3期428-433,共6页
Acta Geodaetica et Cartographica Sinica
基金
国家973计划(2011CB707001)
国家自然科学基金(41102209
41102210
41001308
41071291)
中国博士后科学基金(20110491232)
湖北省自然科学基金(2010CDB04105)
武汉市科技攻关计划(201110821237)
武汉市青年科技晨光计划(201150431074)
中国地质大学(武汉)中央高校基本科研经费(CUG100705
CUG100313)
测绘遥感信息国家重点实验室项目((10)重02)
