论文部分内容阅读
近年来不断发展成熟的合成孔径雷达技术将获取的图像分辨率提高到分米级.在高分辨率条件下,建筑物在SAR图像上表现出的空间信息更加丰富,结构特征更加明显.首先提出了分解模型对高分辨SAR图像中矩形建筑物的特性进行详细分析.在此模型中,散射效应根据不同的贡献来源被细分,以便于解析建筑物图像特征在不同的SAR成像条件下的几何结构和空间分布规律.然后基于建筑物图像表征的结构先验,提出了一种新的单幅高分辨率SAR图像建筑物检测和3-D重建算法,其中包括模型匹配的图像特征的提取,以及先验引导的重建过程.最后,选用了实际高分辨率SAR图像进行建筑物检测和三维重建实验并对重建结果进行了讨论.
In recent years, the mature synthetic aperture radar technology will get the image resolution to the decimeter level.Under the condition of high resolution, the spatial characteristics of the buildings on the SAR images are more abundant and the structural features are more obvious.Firstly, The decomposition model is used to analyze the characteristics of rectangular buildings in high resolution SAR images. In this model, the scattering effects are subdivided according to different contributing sources in order to analyze the geometry of building image features under different SAR imaging conditions Structure and spatial distribution of buildings.Furthermore, based on the structure priori of building image characterization, a new single high resolution SAR image building detection and 3-D reconstruction algorithm is proposed, which includes the extraction of image features of model matching, As well as the a priori guided reconstruction process.Finally, real high-resolution SAR images were selected for building detection and 3D reconstruction experiments and the reconstruction results were discussed.