SAR图像对地定位的严密共线方程模型

SAR Image Localization Using Rigorous SAR Collinearity Equation Model

SAR图像对地定位是解决SAR遥感信息在哪里的根本问题,它是遥感应用和处理的重要步骤。作者在分析现有几种SAR图像对地定位模型的基础上,根据SAR成像的严格几何关系导出SAR图像对地定位的严密共线方程模型。首先根据斜距图像的近地点斜距和远地点斜距推导中心投影等效焦距的计算公式,然后详细推导并给出斜距图像转化成严密中心投影图像时涉及的图像改化公式,最后根据实际星载SAR图像定位的精度对比,证明严密共线方程模型是可行的,而且其定位精度要明显优于传统的近似共线方程模型。

The SAR image localization is the necessary processing procedure before its applications in order to determine the position of SAR image. F. Leberal proposed range-Doppler model to geocode SAR image in 1990 using range equation and Doppler equation of SAR. People engaged in optical remote sensing especially in photogrammetry to geocode the SAR image using collinearity equation in order to understand and process easily. Therefore the geometric models based on collinearity equation have been proposed since 1980s and the basic idea is to transform the SAR range image into central projection image, then the SAR image is geocoded using the optical collinearity equation. This model has been used widely such as G. Konecny. The disadvantage of the model is that approximately processing is always used in the model. A new rigorous SAR collinearity equation model （RSCEM） for SAR image localization is proposed based on rigid geometry of SAR in the paper after analyzing existing models for SAR localization. The equivalent focus length is derived using the rear slant range and far slant range of slant image, and then the formula of correction value to transform slant image into central projection image is given in detail. The new RSCEM model is used to localize the space-borne Radarsat SAR image and the test result shows that positioning accuracy of RSCEM is better than that of the traditional collinearity equation model and better than that range-doppler model.