Determination of 3-dimentional coordinates of Objects from Satellite Images

Abstract

Mathematical framework of the inverse transform from image coordinates to geographic coordinates was developed and investigated. This transform was utilized to estimate the unknown parameters of the satellite imaging system using some of the data extracted from the raw images that produced by the imaging system.A sequence of determinations based on some geometrical relationships between the angular separation of the satellite projection point (nadir points) and a sequence of uniformly distributed reference points were used for autonomous assignment of the initial values of the imaging system parameters. Also, an estimation procedure is given to assess the value of the satellite distance, and an equation for estimating the value of image scale parameter was derived and implemented. Optimization used to improve the determination accuracy. Through each optimization step the value of one of the imaging parameters is adjusted leads to minimize the overall error. The optimization methods have been implemented as two layers. The suggested system was implemented on two sets of Meteosat and Noaa images, and the results justified the suitability of the proposed system.