论文部分内容阅读
由于缺少CBERS-02B星的轨道星历参数和传感器参数,无法利用精确的传感器模型完成CBERS-02B HR数据正射纠正。文章简要介绍了基于平行光投影变换的仿射变换模型原理与方法,选择高山、丘陵、平坝等三种地貌类型的卫星影像数据,利用该模型进行CBERS-02B HR影像高精度几何纠正试验,其纠正结果与多项模型、直接线性变换、有理函数模型纠正的结果相比较,结果表明:利用基于平行光投影变换的仿射变换模型可以完成CBERS-02B HR影像高精度几何纠正,其纠正精度优于常用的纠正模型,可达到国家1∶50000数据库更新工程数字正射影像平面精度的要求。
Due to the lack of orbit ephemeris parameters and sensor parameters of CBERS-02B stars, CBERS-02B HR orthorectification can not be accomplished with an accurate sensor model. In this paper, the principle and method of affine transformation model based on parallel light projection transformation are briefly introduced. The satellite image data of three types of topography, such as alpine, hilly and flat dam, are selected. The model is used to carry out the high accuracy geometric correction test of CBERS-02B HR image. The results of the correction are compared with the results of multiple models, direct linear transformation and rational function model correction. The results show that the high accuracy geometric correction of CBERS-02B HR image can be achieved by affine transformation model based on parallel light projection transformation, Better than the commonly used correction model, can achieve the national 1: 50000 database update project digital orthophoto plane accuracy requirements.