高分辨率SAR图像处理与城市变化监测研究
文献类型：学位论文
| 作者 | 刘斌涛
|
| 学位类别 | 硕士 |
| 答辩日期 | 2009 |
| 授予单位 | 中国科学院水利部成都山地灾害与环境研究所 |
| 授予地点 | 成都 |
| 导师 | 陶和平 |
| 关键词 | 合成孔径雷达 相干斑噪声 图像融合 图像分类 极化目标分解 |
| 其他题名 | Research of High Resolution Synthetic Aperture Radar Image Processing and City Area Land Use Monitoring |
| 学位专业 | 地图学与地理信息系统 |
| 中文摘要 | 雷达遥感是目前遥感中最前沿的技术之一，合成孔径雷达遥感以其独特的全天时、全天候的对地观测能力，备受人们重视。高分辨率合成孔径雷达是雷达遥感重要的发展方向，在近几年得到了飞速的发展。高分辨率SAR图像反应了地物目标丰富的后向散射信息，如何提取和利用这些信息来解决地学中的实际问题都是具有重要意义的研究方向。 本文针对高分辨率星载SAR图像的成像机理和图像特点，深入研究了高分辨率SAR图像的相干斑噪声抑制问题、高分辨率SAR图像融合问题和高分辨图像分类问题，建立了多云多雨区域城市土地利用动态监测技术框架，并提出了可行的技术方法。 本文的主要结论有如下几个方面： (1)提出了基于极化目标分解理论的极化SAR图像与光学多光谱影像融合技术方法，该方法可以将大量的地物散射信息融合进入地物光谱信息中，对提高地物识别具有非常大的意义。 (2)从雷达成像机理和地物散射特性出发，建立了建筑物、道路、水体和森林的目视解译标志和解译方法。通过典型地物对比分析，得出高分辨率SAR图像对人造地物反映非常明显，并主要反映了地物的空间结构信息；在均质区域影响高分辨率SAR图像色调的主要的地物介电常数，主要由水分含量引起；道路和水体同为接近镜面的表面散射，回波很弱，单纯通过图像灰度区分它们比较困难。 (3)通过纹理增强、多极化合成和图像融合的方法都可以提高高分辨率SAR图像的分类精度，应根据实际问题需求和数据条件选择合适的分类策略。通过极化目标分解获得了大量的地物散射特征信息，大幅提高了高分辨率SAR图像的应用潜力。 (4)提出了多云多雨区域城市土地利用变化监测的技术框架，该框架对解决我国西南地区土地利用动态监测和灾害遥感监测都具有非常巨大的实用价值。 |
| 英文摘要 | Radar remote sensing is one of the most cutting-edge technologies in remote sensing areas because it can works in all time and all weather. Synthetic aperture radar image, especially high-resolution synthetic aperture radar image reflects the abundant information of objects on the earth’s surface. So, the methods of improving the image quality for solve the problem in earth science are significative research field. After analyzing the imaging mechanism of synthetic aperture radar and the characteristics of synthetic aperture radar image, the paper studies the high-resolution synthetic aperture radar image speckle suppression, image fusion and image classification. A technical framework had been put forward which can solve the problem of land use monitoring in cloudy and rainy areas using high-resolution synthetic aperture image. The conclusion in this dissertation are as following: (1) This paper put forward a kind of fusion for quad-polarimetric synthetic aperture radar data and multi-spectral image using target decomposition theory. This study can make some earth’s surface’s scattering signature integrated into multi-spectral images, and then more land use classes can be identified from the fused image. (2) From the radar imaging mechanism and scattering characteristics of earth’s surface, the interpret method had been set up, such as buildings, roads, water and forest. There are three results can be obtain from synthetic aperture radar image. First, high-resolution synthetic aperture radar image reflect the structural information of artificial objects, such as buildings and roads. Second, in homogeneous region, the radar image color depends on earth’s surface’s complex permittivity which is caused by soil moisture. Third, it is very difficult to separating road and water if you used gray value only. (3) The high-resolution synthetic aperture radar image classification precision can be improved by texture information enhanced, multi-polarimetric composition and image fusion. We should question the specific needs of the problem and the data conditions for appropriate classification of strategy selection. It is worth mentioning that polarimetric target decomposition is a very effective way to improve the accuracy of synthetic aperture radar image classification. (4) A technical framework which is used to cloudy and rainy areas’ city land use dynamic monitoring was put forward. This technical framework can also use on resources and environment monitoring and natural disaster emergency in southwest China. |
| 学科主题 | 摄影测量与遥感技术 |
| 语种 | 中文 |
| 公开日期 | 2010-10-13 |
| 分类号 | TN9;P20 |
| 源URL | [http://ir.imde.ac.cn/handle/131551/2160]  |
| 专题 | 成都山地灾害与环境研究所_成都山地所知识仓储（2009年以前） 成都山地灾害与环境研究所_数字山地与遥感应用中心 |
| 推荐引用方式 GB/T 7714 | 刘斌涛. 高分辨率SAR图像处理与城市变化监测研究[D]. 成都. 中国科学院水利部成都山地灾害与环境研究所. 2009. |
入库方式： OAI收割
来源：成都山地灾害与环境研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。