高景一号卫星影像DSM自动提取方法

详细信息    查看全文 | 推荐本文 |

英文篇名：Automatic DSM Extraction Based on SuperView-1 Satellite Imagery 作者：陈湘广 ; 张永军 英文作者：CHEN Xiangguang;ZHANG Yongjun;School of Remote Sensing and Information Engineering, Wuhan University; 关键词：高景一号 ; 线阵卫星 ; 核线几何 ; DSM 英文关键词：SuperView-1;;linear array satellite;;epipolarity;;DSM 中文刊名：测绘地理信息 英文刊名：Journal of Geomatics 机构：武汉大学遥感信息工程学院; 出版日期：2019-09-27 出版单位：测绘地理信息 年：2019 期：05 基金：国家重点研发计划(2018YFB0505003) 语种：中文; 页：15-19 页数：5 CN：42-1840/P ISSN：2095-6045 分类号：P237

摘要

利用高景一号数据,首先分析了立体像对核线特性,采用多项式模型生成核线影像;然后,结合影像分块匹配再融合的策略,利用半全局匹配算法进行密集匹配生成视差图;最后通过前方交会实现数字表面模型(digital surface model, DSM)的自动提取。试验结果表明,高景一号影像核线具有平直特性,生成的高分辨DSM宏观地形完整,山地地形重建准确且纹理细节丰富,这表明本文方法应用于高景一号的影像自动提取取得了较好的效果。
        SuperView-1 is China's first ultra-high resolution commercial satellite with spatial resolution up to 0.5 m. Its application marking domestic commercial remote sensing officially steps into the international leading ranks. The stereo data obtained by SuperView-1 is used to analyze, the epipolarity between the stereo pairs, and a polynomial model is used to generate epipolar images. Then, combining the strategy of image block matching and refusion, the semi-global matching algorithm is used for dense matching to generate parallax graph. Finally, the automatic extraction of digital surface model(DSM) is realized through the forward intersection. The experimental results show that the epipolar line of SuperView-1 image has the characteristic of being straight; the high resolution DSM generated by this method has a complete macro terrain; and the mountain terrain is rebuilt accurately and the details are rich in texture. It shows that the application of the proposed method has achieved good results on the SuperView-1.

引文

[1] 郭晗.高景一号正式商用,中国商业遥感进入0.5 m时代[J].卫星应用,2017(5):62-63
    [2] Scharstein D,Szeliski R.A Taxonomy and Evaluation of Dense Two-Frame Stereo Correspondence Algorithms[J].International Journal of Computer Vision,2002,47(1-3):7-42
    [3] Hirschmuller H.Accurate and Efficient Stereo Processing by Semi-Global Matching and Mutual Information[C].IEEE Computer Society Conference on,San Diego,USA,2005
    [4] Haala N.The Landscape of Dense Image Matching Algorithms[C].Photogrammetric Week’13,Berlin,Germany,2013
    [5] 岳庆兴,高小明,唐新明.基于半全局优化的资源三号卫星影像DSM提取方法[J].武汉大学学报·信息科学版,2016,41(10):1 279-1 285
    [6] Zhang L.Automatic Digital Surface Model (DSM) Generation from Linear Array Images[D].Zurich:Swiss Federal Institute of Technology Zurich,2005
    [7] 张云翔,于杰.基于资源三号的DSM自动生成方法与质量控制[J].测绘地理信息,2017,42(4):43-46
    [8] 张静,杨博,王密,等.基于资源3号影像的全国DOM快速制作方法[J].测绘地理信息,2016,41(6):70-74
    [9] 张永军,丁亚洲.基于有理多项式系数的线阵卫星近似核线影像的生成[J].武汉大学学报·信息科学版,2009,34(9):1 068-1 071
    [10] Oh J,Lee W H,Toth C K,et al.A Piecewise Approach to Epipolar Resampling of Pushbroom Satellite Images Based on RPC [J].Photogrammetric Engineering & Remote Sensing,2010,76(12):1 353-1 363
    [11] 胡芬,王密,李德仁,等.基于投影基准面的线阵推扫式卫星立体像对近似核线影像生成方法[J].测绘学报,2009,38(5):428-436