基于极化SAR图像的建筑区提取方法研究

发布时间：2018-04-28 21:14

本文选题：极化合成孔径雷达 + 建筑区提取　；参考：《航天返回与遥感》2017年02期

【摘要】：建筑区提取对于城市规划和灾害评估具有重要的作用。文章提出融合极化方位角补偿的散射模型提取建筑区。首先对极化SAR图像进行极化补偿处理,使得建筑物的散射成分增强,然后融合极化分解得到的散射分量和纹理特征参数作为建筑区提取依据,最后使用面向对象的提取方法进行建筑区提取。采用美国San Francisco地区L波段星载AIRSAR数据和德国Oberpfaffenhofen地区L波段机载ESAR数据分别进行实验验证。结果表明,文中方法较好地识别并提取了建筑区,提高了建筑区的提取精度,可用于受灾建筑区提取以及城市建筑区信息获取。
[Abstract]:Building area extraction plays an important role in urban planning and disaster assessment. In this paper, the scattering model of fusion polarization azimuth compensation is proposed to extract the building area. Firstly, polarization compensation is applied to the polarimetric SAR image to enhance the scattering component of the building, and then the scattering component and the texture characteristic parameters obtained from the polarization decomposition are fused as the basis for the extraction of the building area. Finally, the object-oriented extraction method is used to extract the building area. The L band spaceborne AIRSAR data in San Francisco region of the United States and the L band airborne ESAR data in Oberpfaffenhofen region of Germany are used for experimental verification. The results show that the method can be used to identify and extract the building area, improve the precision of the building area extraction, and can be used to extract the disaster building area and the information of the urban building area.
【作者单位】：南京邮电大学地理与生物信息学院;河海大学地球科学与工程学院;南京烽火软件科技有限公司;
【基金】：南京邮电大学校级科研基金(NY215181);南京邮电大学大学生创新创业训练计划STITP项目(XYB2016273)
【分类号】：TN957.52

【相似文献】