Research On Ghost Image Compression Sensing Algorithm Based On Orthogonal Speckle Computation

Ghost imaging,as a new imaging method that differs from traditional imaging techniques,not only has high resolution during the imaging process,but also has strong anti-interference ability.But early ghost imaging techniques were often not widely used due to their large number of sampling times.Although it has been confirmed in subsequent studies that using ordered orthogonal illumination patterns can effectively shorten imaging time and improve imaging quality,researchers are still exploring solutions to restore images with fewer measurements.This thesis aims to reduce the measurement frequency of ghost imaging and improve the imaging quality of ghost imaging.In order to achieve this research goal,an optimization scheme is proposed based on the use of orthogonal speckle illumination patterns and the selection of appropriate compressed sensing reconstruction algorithm.By comparing ghost imaging under different lighting patterns,it was found that compared with ghost imaging under random lighting patterns,using orthogonal speckle as the lighting pattern for ghost imaging can effectively reduce measurement times and shorten imaging time.This thesis selects two orthogonal illumination patterns: Hadamard speckle and discrete cosine speckle,and conducts numerical simulation and experimental verification of ghost imaging based on different orthogonal speckle calculations.And explored the use of two types of orthogonal speckle patterns for calculating ghost imaging reconstruction results under incomplete measurements,randomly optimized the ghosting problem in Hadamard ghost imaging reconstruction results under incomplete conditions,and further explored the characteristics of the discrete cosine matrix.In order to improve the reconstructed image quality of orthogonal speckle ghost imaging under incomplete conditions,a simple optimization was performed on orthogonal speckle ghost imaging.The numerical simulation and experimental verification of orthogonal speckle ghost imaging based on different compressed sensing algorithms are designed.The research results show that the reconstructed image quality of orthogonal speckle ghost imaging based on OMP algorithm and SP algorithm is lower than that of orthogonal speckle ghost imaging based on TVAL3 algorithm.And through the comparison of image evaluation indicators,it is concluded that the orthogonal speckle ghost imaging based on the TVAL3 algorithm achieves improved imaging efficiency while reducing sampling times.The research results of this thesis provide new insights for improving the imaging quality of ghost imaging,and further lay the foundation for the promotion and use of ghost imaging technology.