Research On Image Restoration Algorithm For Coded Aperture Imaging

The optical imaging detection system is a core system for achieving target detection,recognition detective and battlefield situational awareness.However,due to the limitation of the processing technology of the large-caliber optical lens and the manufacturing process of the detector,the resolution improvement of the conventional optical imaging detection system is difficult.In order to solve this problem the coding aperture technique is applied to the optical imaging system.By adding a specific coding template into the traditional imaging system a large amount of high-frequency information of the object space is entered into the detector,and then the appropriate image reconstruction is performed.The algorithm reconstructs the encoded image to recover the high-frequency information of the target,thereby improving the imaging resolution of the system.However,the existing image reconstruction algorithms have higher requirements on PSF accuracy and have limitations on image detail restoration.This paper studies the code aperture imaging image restoration algorithm,Firstly,the point spread function of the system is estimated,then the improved gradient smoothing boundary algorithm and TVAL3 algorithm are used to recover the blurred image.Finally,experiments were conducted to verify the superiority of the proposed algorithm in image reconstruction.Firstly,the basic principles and models of coded aperture imaging technology are studied.The selection method of optimal coding template,the estimation method of point spread function and the commonly used image quality evaluation methods are studied.Secondly,by using the estimated point spread function and image restoration algorithm for image reconstruction an improved gradient smoothing boundary algorithm is proposed for the ringing effect of Wiener filter restoration algorithm in image reconstruction.The ringing effect can be effectively suppressed.The simulation results show that the proposed algorithm has better suppression effect on the ringing effect than the gradient smoothing boundary algorithm.Thirdly,the full variation restoration algorithm is applied to the image restoration of the coded aperture imaging system,and the commonly used FTVd algorithm and TVAL3 algorithm are studied.These algorithms have lower requirements for the estimation accuracy of the PSF,and the noise can be effectively collected while retaining more image edge with detail information.The simulation results show that the image reconstructed by the total variation algorithm is clear and the resolution is high.Finally,the experimental platform is built,and the coded template in the coded aperture imaging system was screened by KL divergence method and modulation transfer function method.The point spread function of the system at different positions is estimated by the method of pinhole photography,and the point spread function at the position closest to the focus is selected to verify the effectiveness of the coded aperture imaging system;verifies the effectiveness of the proposed algorithm and the various total variation algorithms introduced in this paper.The experimental results show that the improved gradient smoothing boundary algorithm can effectively suppress the ringing effect but can not restore the image details.When the noise is small,the FTVd algorithm has a better restoration effect.When the noise is large,the TVAL3 algorithm has the best restoration effect.