王松^1,2,,
张福博^1,,
陈龙永^1,2,,
梁兴东^1,2,,
①.中国科学院电子学研究所微波成像技术重点实验室 ??北京 ??100190
②.中国科学院大学 ??北京 ??100049
基金项目:国家部委基金

详细信息

作者简介:陈龙永(1979–)，男，研究员，硕士生导师，现任中国科学院电子学研究所微波成像技术重点实验室常务副主任，主要从事高分辨率合成孔径雷达系统、干涉合成孔径雷达系统、微波成像新概念、新体制和新技术等领域的研究工作。E-mail: lychen@mail.ie.ac.cn

通讯作者:梁兴东 ?xdliang@mail.ie.ac.cn

计量

文章访问数:1713
HTML全文浏览量:282
PDF下载量:239
被引次数:0

出版历程

收稿日期:2018-01-19
修回日期:2018-03-20

Array-interferometric Synthetic Aperture Radar Point Cloud Filtering Based on Spatial Clustering Seed Growth Algorithm

Wang Song^1,2,,
Zhang Fubo^1,,
Chen Longyong^1,2,,
Liang Xingdong^1,2,,
①. Science and Technology on Microwave Imaging Laboratory, Institute of Electronics, Chinese Academy of Sciences, Beijing 100190, China
②. University of Chinese Academy of Sciences, Beijing 100049, China
Funds:The National Ministries Foundation


摘要
摘要:阵列干涉合成孔径雷达(Synthetic Aperture Radar, SAR)通过在交轨向布置多个天线，结合方位向的合成孔径和斜距向的大带宽信号，具备了3维分辨能力，且多个阵元保证了其在高程向的空间采样，能够解决干涉SAR(Interferometric SAR, InSAR)测绘中的叠掩问题，实现观测场景的3维成像。但是获得场景区域的3维点云分布中存在较多杂点，高程向误差较大，所以传统的激光雷达(Light Detection And Ranging, LiDAR)点云滤波方法不适用于阵列干涉SAR点云的滤波处理。针对该问题，该文提出基于空间聚类种子生长算法的阵列干涉SAR点云滤波算法，应用密度和高程双重阈值生成密度-高程图像，通过图像处理手段去除小型杂点，利用空间聚类种子生长算法将植被等从点云数据中去除，完成点云滤波处理。利用国内首次机载阵列干涉SAR实验数据，通过与传统LiDAR滤波方法进行比较，验证了该文算法的有效性，为后续建筑物提取和精细化处理提供保障。
关键词:阵列干涉SAR/
点云滤波/
高程图像/
空间聚类种子生长
Abstract:By arranging multiple antennas in the intersection direction and combining the synthetic aperture of azimuth direction and large bandwidth signal with oblique distance, array-interferometric Synthetic Aperture Radar (SAR) can generate a three-dimensional resolution and ensure the elevation spacial sampling due to its multiple array element, which could avoid the layover problem in surveying and mapping in the Interference SAR (InSAR) and realize the three-dimensional imaging of the observation scene. However, considering the existence of too much noise in the three-dimensional point cloud distribution in the scene area and the large elevation error, the traditional Light Detection And Ranging (LiDAR) point cloud filtering method is not suitable for the filtering processing of the array-interferometric SAR point cloud. In order to solve this problem, an array-interferometric SAR point cloud filtering algorithm based on spatial clustering seed growth algorithm is proposed, in which the density-elevation image is generated by the double threshold of density and elevation, the small clutter is removed by image processing, and the vegetation is removed from the point cloud data by using the spatial clustering seed growth algorithm, thus the point cloud filtering process is completed. Using the first airborne array-interferometric SAR experimental data, the validity of the proposed algorithm is verified compared to the traditional LiDAR filtering method, which provides the guarantee for the subsequent building extraction and meticulous treatment.
Key words:Array-interferometric synthetic aperture radar/
Point cloud filtering/
Elevation image/
Spatial clustering seed growth



PDF全文下载地址:

https://plugin.sowise.cn/viewpdf/198_494_R18006