Improved Non-Parametric Geometric Corrections For Satellite Imagery Through Covariance Constraints

摘要: Satellite imagery enables rapid data collection and assessment for several earth environmental sciences. However, the obtained raw cannot be used directly spatial should further processed geometric distortions. The usefulness reliability of from satellite images heavily relies on correction process it can applied through either parametric or non-parametric methods. Parametric methods require physical sensor model parameters which are generally available only to vendor. Rational Polynomial Coefficients (RPCs) supplied by vendors provide global correction, however they usually specific high resolution systems. Other polynomials corrections locally but applicable nearly all Depending degree employed polynomial, topography, distribution quality control points, local could lower misfits at GCPs (Ground Control Points) while might produce very large ICP (Independent Check Points). In particular, when number is limited poorly distributed over image area. this study, 2D 3D Affine functions with covariance constraints introduced improve GCP accuracy distributed. efficiency proposed method was numerically shown in two applications. results show that provides a compromise between without using any coordinates robust efficient lacks precise RPC such as TUBITAK RASAT satellite.