侦测目标的高光谱与SAR图像协同利用与联合分析识别

Hyperspectral remote sensing and synthetic aperture radar are two important detection technologies in the field of geophysical remote sensing. They obtain the physical information of the target from different aspects and play an important role in reconnaissance and anti-reconnaissance. However, due to the image forming mechanism, there are some limitations in both detection technologies. To take full advantage of the two detection technologies, the subject intends to obtain the hyperspectral and microwave scattering data by the indoor and out door synchronization joint observation experiment. And establish a database and a joint simultaneous observation model about the spectral characteristics and scattering properties of the target to reveal the characteristics and its change regularity between the target or camouflage object and natural objects under different observation conditions, and the correlation between hyperspectral image and SAR image. According to the target spectral and scattering characteristics, conduct joint inversion and propose a effective fusion model for the collaborative use and joint analysis of hyperspectral and SAR image. From the research of this paper, we hope it can provide reference value or theoretic foundation and technique measures for the target detection and camouflage evaluation. The innovation of this project is that the research break through the traditional thoughts of image fusion and investigate and reveal the the internal connections and laws between hyperspectral image and SAR image based on the imaging mechanism.

高光谱遥感技术和合成孔径雷达（Synthetic Aperture Radar, SAR）技术是遥感地球物理领域中最重要的两种探测手段，它们从不同的角度获取目标的物性特征信息，在目标侦察与反侦察中发挥着独特的作用。然而，它们的成像机理使其存在各自的局限性。为了使它们优势互补，本课题通过联合观测方式对侦测目标的高光谱及微波散射特性进行室内外同步测量与数据获取，建立地物目标的光谱特性与散射特性实验数据库及联合同步观测模型，揭示目标及伪装物与自然物在不同观测条件下的特性与变化规律及高光谱图像与SAR图像之间的相关关系，根据地物不同的光谱及散射特性，进行联合反演，并提出有效的高光谱和SAR图像协同利用与联合分析融合模型，为侦测目标识别与伪装效果评估提供技术手段和理论依据。本项目的创新之处在于突破传统图像域融合处理思维方式，从两者的成像机理来探讨和揭示它们之间的内在联系和规律。

为了将高光谱遥感技术和合成孔径雷达（Synthetic Aperture Radar, SAR）技术进行优势互补，扬长避短。本课题围绕“侦测目标的高光谱与SAR图像协同利用与联合分析技术”展开研究。首先，通过联合观测方式对侦测目标的高光谱及微波散射特性进行了室内外同步测量与数据获取，并收集了同一地区的高光谱图像与多波段SAR图像。在数据获取的基础上，对目标的高光谱特性以及雷达后向散射特性随成像参数的变化而变化的特性规律进行探索研究，并在此基础上构建了相应的数学模型。最后，重点围绕高光谱与SAR图像的联合分析识别问题，分别对图像预处理、特征提取、图像配准、目标检测等处理方法进行研究，并以此为基础，初步得出一种从多尺度角度出发的多源遥感图像的特征融合与联合识别的技术方法，为侦测目标识别与伪装效果评估提供了新的技术手段和协同利用方法。