直线特征约束的资源3号卫星影像自检校平差

Self-Calibration Adjustment of ZY-3 Satellite Imagery with Linear Feature Constraints

针对控制点获取较困难地区卫星影像定位精度不高的情况,对直线特征作为控制信息提升卫星影像定位精度进行了研究。以"像方直线上任意一点必然位于物方直线和投影中心所构成的平面"作为几何约束条件,通过对直线的参数化表示,建立了基于直线特征的共面模型;在该模型基础上,针对航天传感器的成像特点,分析建立了8标定参数的内方位元素模型和简化的外方位元素模型,最终构建了直线特征约束的卫星影像自检校平差模型。利用资源3号(ZY-3)卫星获取的华盛顿地区数据对构建的平差模型进行实验验证。结果表明,该模型能够解决缺乏地面控制点地区影像定位精度差的问题,可达到与常规自检校平差相同量级的精度。

Considering that some places are tough to obtain ground control points contributing to lower positioning precision of high-resolution satellite imagery,it is studied using linear feature as control feature to improve image positioning accuracy in this paper. Taking＂any point in a straight line of image space having to locate on the plane decided by the corresponding object space straight line and projection center＂as a geometric constraint,the coplanar model based on linear feature is established by parametric expression of linear feature. On the basis of the model,according to the imaging properties of space sensors,interior orientation model and simplified exterior orientation model are set up,finally the self-calibration adjustment model with linear feature constraint is constructed. Using ZY-3 satellite imagery covering Washington area to validate the model,the experimental results show that this model can solve the positioning problem of image-based in lack of ground control points,and it can achieve the equal accuracy compared to normal self-calibration block adjustment.