Optical cryptosystem based on phase-truncated-Fourier-transforms (PTFT) is one of the most interesting optical cryptographic schemes due to its unique mechanism of encryption/decryption. Conventional learning-based attack method need a large number of plaintext-ciphertext pairs to train a neural network and then predict the plaintexts from subsequent ciphertexts. In this work, we propose an alternative method of attack on PTFT-based optical asymmetric cryptosystem by using an untrained neural network. We optimize the parameters of a neural network with the help of the encryption model of PTFT-based cryptosystem, hoping to get the ability of retrieving any plaintext from the corresponding unknown ciphertext but without help of the decryption keys. The proposed untrained-neural-network-based attack approach eliminates the requirement of tens of thousands of training images and might open up a new avenue for optical cryptanalysis.
Scattering light imaging technique has attracted extensive research because of its huge potential in the fields of biomedical microscopy, remote-sensing mapping etc. For most methods now available to reconstruct an object hidden behind scattering media, the main focus is on reconstructing the shape of the object without considering its spectral information. While imaging a color object, it is often necessary to measure a series of Point Spread Functions (PSFs) or Wavelength-Dependent Speckle Patterns (WDSPs) under various wavelengths of illumination. It’s obvious that these methods are either invasive to the object or require multiple exposures. Here, by taking advantage of the Wavelength- Dependent Response Characteristics (WDRC) of the Liquid Crystal Spatial Light Modulator (LC-SLM), we propose an alternative way to reconstruct a hidden color object with noninvasive and single-exposure strategy. A monochromatic camera is adopted to capture the wavelength-multiplexing gray-scale speckle pattern, which can be then demultiplexed into a number of WDSPs by utilizing of a designed Multi-modal Phase Retrieval Algorithm (MM-PRA). Then, a typical speckle correlation technique (SCT) is applied to reconstruct each component of the hidden color object. The feasibility and effectiveness of the proposed method are demonstrated by numerical results in this work while the optical experiments are on the way.
Deconvolution-based techniques have been widely used for imaging through scattering medium due to the optical memory effect (OME) in speckles. Once the point spread function (PSF) of a scattering system is measured, a smallscale object within the OME region can be easily recovered. However, an extended object larger than the OME region can only be partially reconstructed due to the limited field of view (FOV). Here, we find a way to get an integrated PSF by exploiting a point source with different locations in object plane. Thereafter, an extended object, within the FOV but exceeding the OME region, could be recovered by the integrated PSF without knowing any other system parameters even the locations of the point source.