基于不确定性理论的遥感图像几何校正模型优化及定位精度评价方法研究

项目名称： 基于不确定性理论的遥感图像几何校正模型优化及定位精度评价方法研究

项目编号： No.61271013

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 无线电电子学、电信技术

项目作者： 焦伟利

作者单位： 中国科学院遥感与数字地球研究所

项目金额： 60万元

中文摘要： 遥感图像在成像过程中受到各种复杂因素的影响，使其产生几何变形。图像的几何校正实质就是用数学模型来描述物方空间坐标系中的地面点坐标与它在图像平面上像点坐标之间的几何关系，从而消除几何畸变，生成具有地理参考的影像。本项目拟在在分析卫星遥感影像校正过程不确定性的基础上，研究误差传递规律和对模型求解以及校正精度的影响；建立遥感图像校正误差传递模型和修正方案；研究基于多特征的几何校正模型和模型参数选取及优化方法，提高遥感图像校正模型求解精度和稳定性；探讨用不确定性理论分析校正模型的精确性和稳健性，重点研究运用模糊数学理论和灰色系统理论，分析控制点（线、面）精度与分布信息以及其他不确定因素对模型优化的贡献。综合多种不确定因素，研究卫星遥感影像校正模型的定量评价方法和度量指标，进而提出遥感图像几何校正产品空间定位质量的评价方法和可视化表达方法。

中文关键词： 几何校正；不确定性；图像配准；模型优化；定位精度

英文摘要： A variety of complex factors influence the remote sensing image to be geometric position distortion during the image captured. Geometric correction of images in real terms is to use mathematical models to describe the geometric relationship between the features in the image plane and in the object space, thereby eliminating geometric distortion to generate the images with geo-referenced. In this project, we will summarize the law of error propagation and the impact of the errors on the model solving and correction accuracy, based on the analysis of many uncertainties during the processing of image correction. Then we will study on building the error transferring model and error correction schemes for remote sensing image rectification. In particular, we will focus on the research of geometric correction model and its parameters optimization based on the multi-features. We will explore the accuracy and robustness of the calibration model based on the uncertainty theory, especially with fuzzy theory and grey system theory, analyze the influence factors of the model optimization including the accuracy and distribution information of control point (line, area) and other uncertain information. Integrating multifold uncertainties, we will go further research on the quantificational evaluation method and measurable

英文关键词： Geometric Rectification；Uncertainty；Image Registration；Model Optimization；Positioning Accuracy