Imaging through a scattering medium via model-driven deep learning.

Imaging through a scattering medium is of great significance in many areas. Especially, speckle correlation imaging has been valued for its noninvasiveness. In this work, we report a deep learning solution that incorporates the physical model and an additional regularization for high-fidelity speckle correlation imaging. Without large-scale data to train, the physical model and regularization prior provide a correct direction for neural network to precisely reconstruct hidden objects from speckle under different scattering scenarios and noise levels. Experimental results demonstrate that the proposed method presents a significant advance in improving generalization and combating the invasion of noise.

Non-line-of-sight imaging based on an untrained deep decoder network.

In recent years, low-cost high-quality non-line-of-sight (NLOS) imaging by a passive light source has been a significant research dimension. Here, we report a new, to the best of our knowledge, reconstruction method for the well-known "occluder-aided" NLOS imaging configuration based on an untrained deep decoder network. Using the interaction between the neural network and the physical forward model, the network weights can be automatically updated without the need for training data. Completion of the optimization process facilitates high-quality reconstructions of hidden scenes from photographs of a blank wall under high ambient light conditions. Simulations and experiments show the superior performance of the proposed method in terms of the details and the robustness of the reconstructed images. Our method will further promote the practical application of NLOS imaging in real scenes.

Cryptographic analysis on an optical random-phase-encoding cryptosystem for complex targets based on physics-informed learning.

Optical cryptanalysis based on deep learning (DL) has grabbed more and more attention. However, most DL methods are purely data-driven methods, lacking relevant physical priors, resulting in generalization capabilities restrained and limiting practical applications. In this paper, we demonstrate that the double-random phase encoding (DRPE)-based optical cryptosystems are susceptible to preprocessing ciphertext-only attack (pCOA) based on DL strategies, which can achieve high prediction fidelity for complex targets by using only one random phase mask (RPM) for training. After preprocessing the ciphertext information to procure substantial intrinsic information, the physical knowledge DL method based on physical priors is exploited to further learn the statistical invariants in different ciphertexts. As a result, the generalization ability has been significantly improved by increasing the number of training RPMs. This method also breaks the image size limitation of the traditional COA method. Optical experiments demonstrate the feasibility and the effectiveness of the proposed learning-based pCOA method.

Efficient light redirection via stretched field resonance in dielectric meta-resonator.

Light redirection plays an important role in photonic integrated circuit system, which attracts much attention on account of thriving application prospects from microwave to visible frequency. By treating a two-dimensional photonic crystal array as a dielectric resonator with low effective index neff << 1, a new strategy of one-direction semi-enclosed meta-resonator is proposed for light redirection and splitting with a high efficiency beyond 90%. Instead of zero-index material, the phenomenon of significant collimating radiations with zero phase delay can also be achieved through a meta-resonator of low effective index to stretch the internal resonant field with a wavelength much longer than that in air. The geometrical dimensions and structural parameters of the meta-resonator offer a great design flexibility for modulating the operating frequency. The numerical simulation and experimental results perfectly coincide with the theoretical predictions. This strategy can also be extended to other artificial metamaterials and three-dimensional cases, which may offer fantastic possibilities to the development of integrated photonics.

Texture mapping based on photogrammetric reconstruction of the coded markers.

Texture mapping is one of the key procedures to generate photorealistic three-dimensional (3D) models. To avoid dependence on the features of the texture and the geometric model, coded markers are introduced as the control points to assist the texture mapping. Multiple texture images containing the markers are captured, and the 3D coordinates of the markers are reconstructed with photogrammetry; meanwhile, the parameters of the texture camera are optimized with the bundle adjustment strategy. Then the pose parameters of the texture mapping can be calculated with the assistance of the marker registration and the iterative closest point (ICP) algorithms. The validity of the proposed algorithm is demonstrated with an experiment on an ordinary object.

Singular value decomposition ghost imaging.

The singular value decomposition ghost imaging (SVDGI) is proposed to enhance the fidelity of computational ghost imaging (GI) by constructing a measurement matrix using singular value decomposition (SVD) transform. After SVD transform on a random matrix, the non-zero elements of singular value matrix are all made equal to 1.0, then the measurement matrix is acquired by inverse SVD transform. Eventually, the original objects can be reconstructed by multiplying the transposition of the matrix by a series of collected intensity. SVDGI enables the reconstruction of an N-pixel image using much less than N measurements, and perfectly reconstructs original object with N measurements. Both the simulated and the optical experimental results show that SVDGI always costs less time to accomplish better works. Firstly, it is at least ten times faster than GI and differential ghost imaging (DGI), and several orders of magnitude faster than pseudo-inverse ghost imaging (PGI). Secondly, in comparison with GI, the clarity of SVDGI can get sharply improved, and it is more robust than the other three methods so that it yields a clearer image in the noisy environment.

Precise displacement measurement in single-beam interferometry employing photonic metamaterial with effective zero-index.

As interferometry is the highest precision distance measuring technique, we proposed a mechanism of single-beam interferometry employing all-dielectric photonic metamaterial with effective zero-index as a means of precise displacement measurement. This mechanism offers a straightforward method for performing optical range-finding over multi-wavelength and sub-wavelength displacements in a compact uniaxial reflection configuration. The higher sensitivity and resolution can be achieved in this mechanism with the intrinsic accuracy of λ/4. The predesignated measuring device based on the proposed strategy could be directly scaled in dimensions to work at different frequency regions without the need for reconfiguration. Both numerical simulations and experiment have demonstrated its feasibilities and reliability. We believe it will have significant potential applications in the future optical measurement.

Information Security Scheme Based on Computational Temporal Ghost Imaging.

An information security scheme based on computational temporal ghost imaging is proposed. A sequence of independent 2D random binary patterns are used as encryption key to multiply with the 1D data stream. The cipher text is obtained by summing the weighted encryption key. The decryption process can be realized by correlation measurement between the encrypted information and the encryption key. Due to the instinct high-level randomness of the key, the security of this method is greatly guaranteed. The feasibility of this method and robustness against both occlusion and additional noise attacks are discussed with simulation, respectively.

Precise identification of Dirac-like point through a finite photonic crystal square matrix.

The phenomena of the minimum transmittance spectrum or the maximum reflection spectrum located around the Dirac frequency have been observed to demonstrate the 1/L scaling law near the Dirac-like point through the finite ribbon structure. However, so far there is no effective way to identify the Dirac-like point accurately. In this work we provide an effective measurement method to identify the Dirac-like point accurately through a finite photonic crystal square matrix. Based on the Dirac-like dispersion achieved by the accidental degeneracy at the centre of the Brillouin zone of dielectric photonic crystal, both the simulated and experimental results demonstrate that the transmittance spectra through a finite photonic crystal square matrix not only provide the clear evidence for the existence of Dirac-like point but also can be used to identify the precise location of Dirac-like point by the characteristics of sharp cusps embedded in the extremum spectra surrounding the conical singularity.

Triple-image encryption based on phase-truncated Fresnel transform and basic vector operation.

A triple-image encryption method is proposed that is based on phase-truncated Fresnel transform (PTFT), basic vector composition, and XOR operation. In the encryption process, two random phase masks, with one each placed at the input plane and the transform plane, are generated by basic vector resolution operations over the first and the second plaintext images, and then a ciphered image in the input plane is fabricated by XOR encoding for the third plaintext image. When the cryptosystem is illuminated by an on-axis plane, assisted by PTFT, the ciphered image is finally encrypted into an amplitude-only noise-like image in the output plane. During decryption, possessing the correct private key, decryption keys, and the assistant geometrical parameter keys, and placing them at the corresponding correct positions, the original three plaintext images can be successfully decrypted by inverse PTFT, basic vector composition, and XOR decoding. Theoretical analysis and numerical simulations both verify the feasibility of the proposed method.

Multiple-image authentication with a cascaded multilevel architecture based on amplitude field random sampling and phase information multiplexing.

A multiple-image authentication method with a cascaded multilevel architecture in the Fresnel domain is proposed, in which a synthetic encoded complex amplitude is first fabricated, and its real amplitude component is generated by iterative amplitude encoding, random sampling, and space multiplexing for the low-level certification images, while the phase component of the synthetic encoded complex amplitude is constructed by iterative phase information encoding and multiplexing for the high-level certification images. Then the synthetic encoded complex amplitude is iteratively encoded into two phase-type ciphertexts located in two different transform planes. During high-level authentication, when the two phase-type ciphertexts and the high-level decryption key are presented to the system and then the Fresnel transform is carried out, a meaningful image with good quality and a high correlation coefficient with the original certification image can be recovered in the output plane. Similar to the procedure of high-level authentication, in the case of low-level authentication with the aid of a low-level decryption key, no significant or meaningful information is retrieved, but it can result in a remarkable peak output in the nonlinear correlation coefficient of the output image and the corresponding original certification image. Therefore, the method realizes different levels of accessibility to the original certification image for different authority levels with the same cascaded multilevel architecture.

Optical identity authentication scheme based on elliptic curve digital signature algorithm and phase retrieval algorithm.

An optical identity authentication scheme based on the elliptic curve digital signature algorithm (ECDSA) and phase retrieval algorithm (PRA) is proposed. In this scheme, a user's certification image and the quick response code of the user identity's keyed-hash message authentication code (HMAC) with added noise, serving as the amplitude and phase restriction, respectively, are digitally encoded into two phase keys using a PRA in the Fresnel domain. During the authentication process, when the two phase keys are presented to the system and illuminated by a plane wave of correct wavelength, an output image is generated in the output plane. By identifying whether there is a match between the amplitude of the output image and all the certification images pre-stored in the database, the system can thus accomplish a first-level verification. After the confirmation of first-level verification, the ECDSA signature is decoded from the phase part of the output image and verified to allege whether the user's identity is legal or not. Moreover, the introduction of HMAC makes it almost impossible to forge the signature and hence the phase keys thanks to the HMAC's irreversible property. Theoretical analysis and numerical simulations both validate the feasibility of our proposed scheme.

Digital color image watermarking based on phase-shifting interferometry and neighboring pixel value subtraction algorithm in the discrete-cosine-transform domain.

A novel single-channel color-image watermarking with digital-optics means based on phase-shifting interferometry (PSI) and a neighboring pixel value subtraction algorithm in the discrete-cosine-transform (DCT) domain is proposed. The converted two-dimensional indexed image matrix from an original color image is encrypted to four interferograms by a PSI and double random-phase encoding technique. Then the interferograms are embedded in one chosen channel of an enlarged color host image in the DCT domain. The hidden color image can be retrieved by DCT, the improved neighboring pixel value subtraction algorithm, an inverse encryption process, and color image format conversion. The feasibility of this method and its robustness against some types of distortion and attacks from the superposed image with different weighting factors are verified and analyzed by computer simulations. This approach can avoid the cross-talk noise due to direct information superposition, enhance the imperceptibility of hidden data, and improve the efficiency of data transmission.

Information security system by iterative multiple-phase retrieval and pixel random permutation.

A novel information security system based on multiple-phase retrieval by an iterative Fresnel-transform algorithm and pixel random permutation (PRP) technique is proposed. In this method a series of phase masks cascaded in free space are employed and the phase distributions of all the masks are adjusted simultaneously in each iteration. It can achieve faster convergence and better quality of the recovered image compared with double-phase encoding and a similar approach in the spatial-frequency domain with the same number of phase masks and can provide a higher degree of freedom in key space with more geometric parameters as supplementary keys. Furthermore, the security level of this method is greatly improved by the introduction of the PRP technique. The feasibility of this method and its robustness against occlusion and additional noise attacks are verified by computer simulations. The performance of this technique for different numbers of phase masks and quantized phase levels is investigated systematically with the correlation coefficient and mean square error as convergence criterions.

Phase-only encryption and watermarking based on phase-shifting interferometry.

Generally, the reconstruction of an object image from its diffraction field requires both the amplitude and the phase information of this field. We systematically investigated the effects of using only the real part, the imaginary part, or the phase information of the diffraction field to reconstruct the original image for both the binary and the gray-level images. We show that the phase information can yield a better result of image retrieval than the real or imaginary part and that the recovered image from the phase information is satisfactory especially for binary input. On the basis of this idea, a new technique of image encryption and watermarking by use of only one delivered image-the phase map of the diffraction field of the original image-through double random-phase encoding is proposed and verified by computer simulations with phase-shifting interferometry. This method can greatly cut down the communication load and is suitable for Internet transmission.

Digital image encryption and watermarking by phase-shifting interferometry.

A method for both image encryption and watermarking by three-step phase-shifting interferometry is proposed. The image to be hidden is stored in three interferograms and then can be reconstructed by use of one random phase mask, several specific geometric parameters, and a certain algorithm. To further increase the security of the hidden image and confuse unauthorized receivers, images with the same or different content can be added to the interferograms, and these images will have no or only a small effect on the retrieval of the hidden image, owing to the specific property of this algorithm. All these features and the utility of this method for image retrieval from parts of interferograms are verified by computer simulations. This technique uses intensity maps as decrypted images for delivery, and both encryption and decryption can be conveniently achieved digitally. It is particularly suitable for the remote transmission of secret information via the Internet.

Polarization optimization in the interference of four umbrellalike symmetric beams for making three-dimensional periodic microstructures.

A systematic and comprehensive analysis of the interference of four umbrellalike beams (lFUB) is provided based on the reciprocal space theory. The concept of pattern contrast is extended to the case of the IFUB, and it is indicated that a uniform contrast for all the interference terms can be obtained by properly choosing the beam ratio and the polarization of each beam. Different polarization combinations, including linear light and linear light, circular light and circular light, and linear light and circular light, have been discussed for the purpose of maximum uniform contrast. It is shown that the use of circular light may generally improve the uniform contrast. This study may lay a theoretical foundation for holographic fabrication of three-dimensional (3D) periodic microstructures, such as simple cubic, body-centered cubic, face-centered cubic, or trigonal lattice.