Application Research Of Real-time Electron Racemization Technology On Strapdown Rolling Platform

Strapdown imaging platform can save a lot of space and cost for TV terminal guided weapons and improve the reliability of the system.However,during the rolling motion of the carrier,the imaging equipment attached to the carrier also moves with it,which makes the image superimpose additional rotation and even blur the image.The image electronic elimination technology can solve the image rotation problem by only using some algorithm to process the image at the pixel level,so as to facilitate the subsequent observation and processing of the image.It has wide application space in both military and civil fields.Based on the moving characteristics of image terminal guided weapons,this paper points out the image rotation and rotation motion blur problems of strapdown imaging platform,and this paper focus on the image rotation,interpolation on embedded equipment and rotation motion blur restoration research.First of all,in order to improve the image quality of real-time electron rotation elimination algorithm,aiming at the problem that the bilinear interpolation is not good at the edge and the gradient interpolation requires a large amount of computation,a sectional interpolation algorithm based on image edge is designed to protect the image edge while giving consideration to the real-time performance of the algorithm.Aiming at the error problem of traditional edge detection algorithm in non-integer coordinates,this paper designs an image edge detection method based on the four neighborhoods' first order differentials of the points to be interpolated,and uses the local gradient feature of the image to suppress the non-maximum value of the edges,and obtains a continuous and accurate edge detection algorithm for non-integer coordinates pixels.The simulation results show that the image edges obtained by sectional interpolation with the image edges found by our algorithm in this paper are prominent and the non-edge areas are smooth,which improve the image quality of the electronic elimination system while considerate the real-time performance.Secondly,aiming at the problem of large computation of compensated image rotation and deblurring algorithm,we design a fast extraction method of pixels along the rotation path,the image rotation problem can be converted to the pixel sequences moving along the rotation path,rotation motion blur problems can be converted to the one dimensional motion blur along the rotation path,this algorithm can be combined with the sectional interpolation algorithm could achieve the image rotation elimination by only use some simple arithmetic and logical operations.The error analysis shows that the maximum relative error of the rotation path found by the method in this paper is less than 5.355%,and the larger the radius,the smaller the error,which meets the practical application requirements.At the last,this paper analyzes the performance of the relevant algorithms.The image rotation algorithm is stable at all angles,and the rotation motion blur restoration algorithm is obviously better than the traditional method in the large-scale blur restoration.When the blur scale is 30°,the PSNR of this paper's algorithm is 1.87 db higher than that of the polar coordinate method.In the equivalent dynamic experiment,the proposed algorithm achieves high quality image rotation elimination and verifies the feasibility of the proposed algorithm.