Speckle autocorrelation imaging based on the parallel amplitude-phase retrieval algorithm

In scattering imaging, a light beam carrying target information will form a speckle image after passing through a scattering medium. In order to recover the original target image from the speckle image, this paper designs an optical imaging scheme of a light beam passing through a random scattering medium, based on the optical memory effect and speckle autocorrelation technology. This paper proposes a novel reconstruction algorithm based on multi-range coherent diffraction imaging, named parallel Amplitude-Phase Retrieval (APR) algorithm. The simulation is carried out under the condition of static scattering medium. The performance of the reconstructed images of different restoration algorithms is compared by the Peak Signal to Noise Rate (PSNR). Through the analysis of the simulation results, for the reconstruction of simple images, the PSNR of APR algorithm is 35.9 dB, which is 6.6 dB, 7 dB and 6.5 dB higher than that of ER algorithm, HIO algorithm and ER-HIO algorithm respectively. For complex image reconstruction, the PSNR of APR algorithm is 26.3 dB, which is 8 dB, 7.2 dB and 6.5 dB higher than that of ER algorithm, HIO algorithm and ERHIO algorithm respectively. The results show that APR algorithm is better than other algorithms for image reconstruction.