Correction Of Registration Errors In Solar High Resolution Sequence Images

Earth and space telescopes produce a large number of sun high resolution images,in order to better obtain the dynamic information in the sun image,the need for sequential image processing research,rather than confined to a single frame image.In the process of registering the high-resolution sequence image of the sun,there will be some systematic difference.For the whole sequence image,the smaller error between the two images will cause the drift of the whole sequence image due to the accumulation of the error.In this paper,we propose a method of adjusting the effect of accumulating and accidental error of system error after the whole sequence image registration in the process of registering the whole sequence image.In this paper,the simulated image and the real image data are analyzed experimentally.In the simulation experiment,a set of sequence images is constructed first,and the displacement values between the images are obtained by registering the sequence images.According to the relationship between the displacement value of the image,the displacement value is added with the corresponding correction number to construct the error equation.According to the least squares principle and the Lagrangian multiplier method,the equation is constructed.Solve the equation and get the value of the corrected number.The correction value is obtained by adding the corresponding value to the displacement value.In the simulation test,different strategies were tested,and in the process of processing the sequence image,several images were re-selected on the selected strategy for correction.As well as the image of the noise,deformation and other treatment.Through different tests,it is shown that the adjustment method is effective for the correction of the sequence image registration error.In the process of correcting the solar high-resolution sequence of SOT,the theoretical value of the sequence image can not be simulated as the reference of the final correction value because it can not simulate the experiment.By registering different areas of the same set of sequence images.Comparing the corrections with different measured values to obtain a corresponding set of correction values.Finally,the measured values and correction values of each group were analyzed on the results graph.Also confirmed the effectiveness of the adjustment method..