| 机载LiDAR点云时间纹理信息航带重叠区消冗 |
| |
| 引用本文： | 邹长江,马洪超,张良,巩翼龙,刘清川.机载LiDAR点云时间纹理信息航带重叠区消冗[J].遥感学报,2014,18(6):1208-1216. |
| |
| 作者姓名： | 邹长江 马洪超 张良 巩翼龙 刘清川 |
| |
| 作者单位： | 武汉大学 遥感信息工程学院, 湖北 武汉 430079;武汉大学 遥感信息工程学院, 湖北 武汉 430079;武汉大学 遥感信息工程学院, 湖北 武汉 430079;武汉大学 测绘信息工程学院, 湖北 武汉 430079;武汉大学 遥感信息工程学院, 湖北 武汉 430079 |
| |
| 基金项目： | 国家自然科学基金(编号:61378078);国家科技支撑项目(编号:2012BAH34B02);武汉大学研究生自主科研项目(编号:2012213020203) |
| |
| 摘 要： | 机载LiDAR获取的完整测区点云数据中包含了丰富的信息,同时也伴随着大量冗余数据,本文提出基于机载LiDAR点云时间纹理信息的航带重叠区消冗方法。首先按点云时间信息划分航带,再按点云纹理信息提取航带边缘,接着提取高地物遮挡空洞,最后去除重叠区冗余数据。实验结果表明,该方法无需航线信息辅助,并能在保留遮挡空洞区域点云的同时,高效地去除航带重叠区中精度较低的点云。
|
| 关 键 词： | 机载LiDAR 时间纹理信息 航带重叠区 遮挡空洞 数据消冗 |
| 收稿时间： | 2014/3/17 0:00:00 |
| 修稿时间： | 2014/6/9 0:00:00 |
Data reduction in overlapping airborne LiDAR strips based on time and texture information |
| |
| ZOU Changjiang,MA Hongchao,ZHANG Liang,GONG Yilong and LIU Qingchuan.Data reduction in overlapping airborne LiDAR strips based on time and texture information[J].Journal of Remote Sensing,2014,18(6):1208-1216. |
| |
| Authors: | ZOU Changjiang MA Hongchao ZHANG Liang GONG Yilong and LIU Qingchuan |
| |
| Institution: | School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China;School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China;School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China;School of Geodesy and Geomatrics, Wuhan University, Wuhan 430079, China;School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China |
| |
| Abstract: | Redundant point clouds are inevitable in the overlapping area of adjacent airborne LiDAR strips. Data redundancy causes problems in terms of data density and accuracy consistency compared with other regions. Reduction of airborne LiDAR data is an important step in LiDAR data pre-processing to distribute the laser points more evenly and eliminate accuracy inconsistency. This paper aims to reduce point cloud without auxiliary data. An algorithm based on time and texture information extracted from point cloud is proposed.(1)Point clouds are partitioned into different strips using cluster analysis method according to their GPS time.(2)A texture graph is generated by analyzing the LiDAR scanning mode and projecting point coordinates to scanning detection, from which the key points are extracted and connected into strip edge lines.(3)When holes caused by the occlusion of high objects appear: Ground and non-ground points are separated automatically by adaptive TIN filtering, followed by classifying high object points based on region-growing method, and finally detecting the edges of the high objects by Alpha-Shapes algorithm. After combining with other the information contained in point clouds, the holes occluded by high objects can be detected.(4)Remove of the redundant point clouds with low accuracy in the overlapping area by comparing the distances of the points to their strip edge lines. Thus, the closer distance points from their strip edge lines, the lower accuracy these points have. The redundant and low accuracy point clouds in the overlapping area were efficiently removed and the hole areas were reserved. Meanwhile, the accuracy of the DEM generated from the simplified data slightly improved. Experiments results showed the effectiveness of our method, which outperformed the conventional data reduction approaches in terms of the fact that flight trajectory information is necessary in those methods. |
| |
| Keywords: | airborne LiDAR time and texture information overlapping strip area shadow hole data simplification |
|
| 点击此处可从《遥感学报》浏览原始摘要信息 |
| 点击此处可从《遥感学报》下载免费的PDF全文 |
|
