The satellite attitude calculation method for linear CCD imaging the same region of non-uniform scene was presented, and a nonlinear relative radiometric calibration method based on histogram matching was put forward. Firstly, when it was needed to perform relative radiometric calibration task, the initial drift angle was calculated, according to which the yaw angle could be adjusted to ensure on-orbit satellite performing the calibration imaging mode. Secondly, in calibration imaging process the yaw angle was controlled to ensure all CCD detectors imaging the same region. Finally, after obtaining the same region image with CCD camera, histogram matching method was used to establish high-precision nonlinear relative radiometric calibration model. The simulation results show the magnitude and time interval of yaw angle adjustment with various attitudes, such as nadir imaging, roll attitude, pitch attitude, roll and pitch attitude, etc, and also analyze the error factors and the uncertainty of drift angle. This method needs neither the ground uniform scaling field, nor lots of on-orbit images for statistical analysis. More importantly, each track satellite images can meet the calibration condition, which avoids the unreliable problem of calibration source itself caused by the unstable of images between different tracks.