An imaging geometry model of space camera

Abstract

High resolution of remote sensing image is impressible on varying pitch angle of satellite platform on orbit. The geometry quality of image is distorted, and image has geometrical warp. Consequently, the spatial distribution of image is changed. However, traditional simulation methods of geometric distortion are complex. Traditional methods are based on accurate physical model. The pixel positions of warp image are calculated as one by one pixel. The topological mapping relationship is analyzed, which is between earth coordinate and optical remote sensor coordinate. The method of active points is proposed. Positions of active points are computed through the transform relationship between earth coordinate and optical remote sensor coordinate. Active points are interpolated by polynomial interpolation. The geometrical distortion is sub-pixel precision. Finally, a frame of image is generated. The effective transform reduces vastly amount of computation. The geometry model contains interior and exterior orientation elements of imaging system on satellite platform. The simulation experiment is based on three axes. Various angles of three axes are included by proposed model. As a result, the boundary condition of motion error affecting imaging quality is analyzed. The proposed geometry model not only improves physical information of active points, but also reduces computational complexity of transform between earth coordinate and optical remote sensor coordinate. The result is beneficial to design and optimize parameters of satellite platform.