Fundamental Analytics of Satellite Imageries

点击翻译按钮获取中文摘要

The orientation problem of satellite imageries such as satellite CCD line-scanner imagery and satellite CCD camera imagery was fundamentally investigated in this research. First, orientation theories of affine imageries was newly constructed for both one-dimensional imagery and two-dimensional one, and the geometrical characteristics were clarified for the relative orientation process (the model construction phase) and the absolute orientation process (the one-to-one correspondence relating the model and object spaces). Then, the transformation of central-perspective imagery into affine one was explored, because we have in reality no technique to take affine imageries of objects directly. The developed orientation approaches of satellite imageries were tested with many simulated examples so as to clarify the difficulties when applying them to the practical case. Also, the derived orientation theory of two-dimensional affine imagery is revealed to be very effectively applicable for not only the analysis of zoom-lens CCD camera imagery but ultra-precise measurement of small objects, which are very important in engineering photogrammetry.Studies on the orientation problem of overlapped radar imageries revealed that:(1) the model construction with overlapped imageries based on ranging principle are also possible,(2) the one-to-one correspondence between the model and object spaces can be uniquely determined with eight independent parameters, and(3) the self-calibration of a radar scanner is also possible without object space information.