Modeling And Analysis Of Image Motion Velocity For Satellite Borne TDICCD Camera In Roll Attitude

Imaging with high resolution in the wide Field-of-view (FOV) is the main task ofspace borne scanning TDICCD remote sensing camera. Precise compensation of imagemotion is the precondition of the camera providing high quality images. How to enlargethe FOV of the scanning system and build exact model of image motion is the primarywork. This article models the image motion velocity of side-sway scanning based on thesub-satellite point image motion velocity model of the space borne TDICCD camera.The velocity error and the drift angle error of the image motion are analyzed. The maintasks are:Firstly, the method of building sub-satellite point image motion velocity model ofthe space borne TDICCD camera was deeply analyzed, using homogeneous coordinatetransformation. The transformation relation between each coordinate system and thefunction of each parameter are studied. The parameters in the image motion velocitymodel were definitely defined, laying the theoretical basis for building side-swayscanning model of image motion.Secondly,45-degree-rotated scanning mirror (RS Mirror) was used to expandscanning field of the system. According to the imaging character of the RS Mirror, weused K mirror (the optical system is composed of three plane mirrors and the shape likes‘K’ word, so it’s named ‘K mirror’) to eliminate the image rotation and proved itsrationality by simulation. The transformation matrices of the RS Mirror and the Kmirror were deduced. We also build the side-sway scanning model of image motionvelocity based on the sub-satellite point image motion model. Parameters, like i’0, γ’0and h(), were redefined and their changes were analyzed when the satellite was indifferent positions of the orbit.Then, we analyzed changes of the image motion speed of each image point inforward and lateral direction and the drift angle when the satellite was in two situations:the same side-sway angle but different latitudes and the same latitudes but differentside-sway angles. The result showed that the image rotation angle was the main factorthat affects the image motion speed and the drift angle when using detector rotation torevise image rotation angle and the drift angle.At last, according to the relationship between the relative error of the image motionspeed and the MTF, we provided limit value of the needed relative error of the imagemotion speed the drift angle error. The changing curve of the relative matching error ofthe image motion speed (EVp2) and the compensating errors of the drift angle (Eη2) weresimulated when the satellite was in two side-sway imaging methods and twocompensating schemes. Combining the two limit values, we derived limit situation of the different side-sway imaging. We also analyzed the estimation error of the imagemotion speed (ΔVp) and the drift angle (Δηp) in the same side-sway imaging by usingMonte Carlo method. It is proved the relative matching error of the image motion speed(EVp2) and the compensating errors of the drift angle (Eη2) were main characters thataffect the image quality.