SAR图像中建筑物模型分析与重建

发布时间：2018-03-28 16:46

本文选题：合成孔径雷达　切入点：建筑物模型分解　出处：《西安电子科技大学》2014年硕士论文

【摘要】：合成孔径雷达(SAR)是一种以小孔径天线利用脉冲技术,合成孔径原理实现的具有高分辨率的新型雷达系统,较之原有的雷达系统,其功能得到质的变化。SAR可以对局域目标进行成像识别,同时在遥感应用方面表现出巨大的潜能,给人们提供更多的信息帮助人们探索地表目标。SAR具有全天候,远距离,高分辨率等优点,对国家的国防信息化建设,国民经济发展和科技进步具有重要的意义,因此,SAR技术受到越来越多国家的重视和青睐,是当前一个快速发展的技术领域。本文对SAR成像系统、SAR成像原理及特点、SAR图像的多种特性以及SAR图像中常见建筑物散射效果做了详细的介绍。并根据SAR图像中矩形建筑物的散射特性,给出了矩形建筑物的分解模型,用以具体阐释SAR图像中矩形建筑物的特点。同时给出了用以建立目标几何空间维度信息和图像特征之间关系的数学公式,将模糊的问题转化为详细的空间几何结构关系对图像的特征进行提取。然后,根据矩形建筑物模型,给出了一种新的检测和重建算法,包括基于匹配度算法的特征提取以及建筑物重建过程。但是,该方法的使用环境仍然有着很多的局限性,在多散射效果的情况下、复杂噪声的影响以及相邻对象的相互影响与遮挡的情况下,仍然不能很好对目标建筑物进行三维重建,更先进以及更好的检测和重建算法将值得进一步的研究,以应用在更复杂的场景。
[Abstract]:Synthetic Aperture Radar (SAR) is a new type of radar system with high resolution, which is realized by using pulse technology and synthetic aperture principle with small aperture antenna, compared with the original radar system. Its function can be changed qualitatively. SAR can be used to image and identify local targets. At the same time, it shows great potential in remote sensing applications, and provides people with more information to help people explore the surface targets. The advantages of high resolution are of great significance to the national defense information construction, the development of national economy and the progress of science and technology. Therefore, the SAR technology has been paid more and more attention and favor by more and more countries. In this paper, the principle and characteristics of SAR imaging system are introduced in detail, as well as the scattering effects of buildings in SAR images. According to the moments in SAR images, this paper gives a detailed introduction to the principle and characteristics of SAR imaging system. Scattering characteristics of a building, The decomposition model of rectangular building is given to explain the characteristics of rectangular building in SAR image, and the mathematical formula is given to establish the relationship between dimension information of object geometry space and image feature. The fuzzy problem is transformed into a detailed spatial geometric structure relation to extract the features of the image. Then, according to the rectangular building model, a new detection and reconstruction algorithm is proposed. This method includes feature extraction based on matching degree algorithm and building reconstruction process. However, the environment of this method still has many limitations, in the case of multi-scattering effect, Under the influence of complex noise and the interaction and occlusion of adjacent objects, the 3D reconstruction of the target buildings is still not good. More advanced and better detection and reconstruction algorithms will be worthy of further study. To apply in more complex scenarios.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TU198;TN957.52

【相似文献】