Serial GANs: A Feature-Preserving Heterogeneous Remote Sensing Image Transformation Model

In recent years, the interpretation of SAR images has been significantly improved with the development of deep learning technology, and using conditional generative adversarial nets (CGANs) for SAR-to-optical transformation, also known as image translation, has become popular. Most of the existing image translation methods based on conditional generative adversarial nets are modified based on CycleGAN and pix2pix, focusing on style transformation in practice. In addition, SAR images and optical images are characterized by heterogeneous features and large spectral differences, leading to problems such as incomplete image details and spectral distortion in the heterogeneous transformation of SAR images in urban or semiurban areas and with complex terrain. Aiming to solve the problems of SAR-to-optical transformation, Serial GANs, a feature-preserving heterogeneous remote sensing image transformation model, is proposed in this paper for the first time. This model uses the Serial Despeckling GAN and Colorization GAN to complete the SAR-to-optical transformation. Despeckling GAN transforms the SAR images into optical gray images, retaining the texture details and semantic information. Colorization GAN transforms the optical gray images obtained in the first step into optical color images and keeps the structural features unchanged. The model proposed in this paper provides a new idea for heterogeneous image transformation. Through decoupling network design, structural detail information and spectral information are relatively independent in the process of heterogeneous transformation, thereby enhancing the detail information of the generated optical images and reducing its spectral distortion. Using SEN-2 satellite images as the reference, this paper compares the degree of similarity between the images generated by different models and the reference, and the results revealed that the proposed model has obvious advantages in feature reconstruction and the economical volume of the parameters. It also showed that Serial GANs have great potential in decoupling image transformation.

All-optical synthesis of an arbitrary linear transformation using diffractive surfaces

Spatially-engineered diffractive surfaces have emerged as a powerful framework to control light-matter interactions for statistical inference and the design of task-specific optical components. Here, we report the design of diffractive surfaces to all-optically perform arbitrary complex-valued linear transformations between an input (Ni) and output (No), where Ni and No represent the number of pixels at the input and output fields-of-view (FOVs), respectively. First, we consider a single diffractive surface and use a matrix pseudoinverse-based method to determine the complex-valued transmission coefficients of the diffractive features/neurons to all-optically perform a desired/target linear transformation. In addition to this data-free design approach, we also consider a deep learning-based design method to optimize the transmission coefficients of diffractive surfaces by using examples of input/output fields corresponding to the target transformation. We compared the all-optical transformation errors and diffraction efficiencies achieved using data-free designs as well as data-driven (deep learning-based) diffractive designs to all-optically perform (i) arbitrarily-chosen complex-valued transformations including unitary, nonunitary, and noninvertible transforms, (ii) 2D discrete Fourier transformation, (iii) arbitrary 2D permutation operations, and (iv) high-pass filtered coherent imaging. Our analyses reveal that if the total number (N) of spatially-engineered diffractive features/neurons is ≥Ni × No, both design methods succeed in all-optical implementation of the target transformation, achieving negligible error. However, compared to data-free designs, deep learning-based diffractive designs are found to achieve significantly larger diffraction efficiencies for a given N and their all-optical transformations are more accurate for N < Ni × No. These conclusions are generally applicable to various optical processors that employ spatially-engineered diffractive surfaces.

Reduction of FDTD Stair-Casing Error regarding to Metamaterials by Using High-Order Polynomial Transformation Functions

The material parameters of a metamaterial (MTM) are determined by the transformation function used in the optical transformation. Some previously reported MTMs, such as the invisibility cloak, the field rotator, and the field concentrator, were designed by a linear transformation. Their impedance was matched to the background so that no reflection was found; however, the material parameters were mismatched to the background due to the linear transformation function. In the present work, the parameters were matched by using high-order polynomial functions as the transformation function. Since similar materials are filled in boundary cells of the finite-difference time-domain (FDTD) algorithm, the stair-casing error was reduced and the tolerance against boundary abrasion was increased. The frequency response of the proposed method was analyzed. The proposed method is applicable to MTM structures that have complex boundary shapes. In this work, circular and elliptic boundary shapes were considered as examples.

Some problems of modeling laser-induced filaments

Experimental data of formation laser-induced filaments in various media (potassium chloride, water, air and silicon carbide) are represented. These phenomena are analyzed as processes of Nonlinear and Relaxed Optics. Problems of modeling the creation the volume laser-induced filaments are investigated. Comparative analysis of plasma, nonlinear optical, diffractive and interference phenomena (including diffractive stratification), shocking processes (including Cherenkov radiation) and physical-chemical processes (including cascade model of excitation the proper chemical bonds in the regime of saturation the excitation), methods and models are represented and discussed. The optical breakdown for various matters has various natures: from shock ionization of gas to disruption of all chemical bonds for solid in the region of interaction light and matter or from nonequilibrium radiated processes in gas and liquid to irreversible phase transformations in solid. For diffraction stratification the modified models of Rayleygh rings was used. We show that this model allow to explain the experimental data for silicon carbide more effectively as Lugovoy-Prokhorov theory of moving foci. Modified Niels and Aage Bohrs models (microscopic) and Golub model (macroscopic) of Cherenkov radiation were used for the explanation of generation continuous radiation. Diffraction stratification shows the surface conic nature of Cherenkov radiation. It was show that physical-chemical method of estimations of corresponding processes is more general as electromagnetic (Kerr media) and one allow explain basic terms of resulting chain process with united point of view. Modified I. Frank model of interference the Cherenkov radiation was used for the explanation laser-induced optical breakdown in silicon carbide. Modified Rayleygh model and methods of continuum mechanics was created and used for the estimation sizes and form of observing nanovoids of silicon carbide. In whole the represented models allow to explain the corresponding chain more fuller and really as other models because one take into account of nonlinear optical transformation of primary laser radiation.

Mechanoluminescence of a composite based on polymethylmethacrylate and fine-disperse powder of phosphor excited by short acoustic pulses and dynamic action of the stylus

A composite mechanoluminescent layer has been produced on the surface of polymethylmethacrylate by liquid-phase embedding of [Formula: see text] phosphor microparticles into the polymethylmethacrylate surface layer. The photoluminescence and mechanoluminescence of the obtained layer have been investigated. The mechanoluminescence was excited by the short acoustic pulses and by the dynamic pressure of the stylus sliding over the mechanoluminescent layer surface. A possible mechanism of mechanoluminescence excitation is under discussion. The produced composite layer is shown to exhibit high efficiency of “mechano-optical” transformation.

Multiplication and division of the orbital angular momentum of light with diffractive transformation optics

We present a method to efficiently multiply or divide the orbital angular momentum (OAM) of light beams using a sequence of two optical elements. The key element is represented by an optical transformation mapping the azimuthal phase gradient of the input OAM beam onto a circular sector. By combining multiple circular-sector transformations into a single optical element, it is possible to multiply the value of the input OAM state by splitting and mapping the phase onto complementary circular sectors. Conversely, by combining multiple inverse transformations, the division of the initial OAM value is achievable by mapping distinct complementary circular sectors of the input beam into an equal number of circular phase gradients. Optical elements have been fabricated in the form of phase-only diffractive optics with high-resolution electron-beam lithography. Optical tests confirm the capability of the multiplier optics to perform integer multiplication of the input OAM, whereas the designed dividers are demonstrated to correctly split up the input beam into a complementary set of OAM beams. These elements can find applications for the multiplicative generation of higher-order OAM modes, optical information processing based on OAM beam transmission, and optical routing/switching in telecom.

A Hybrid Bio-Optical Transformation for Satellite Bathymetry Modeling Using Sentinel-2 Imagery

The article presents a new hybrid bio-optical transformation (HBT) method for the rapid modelling of bathymetry in coastal areas. The proposed approach exploits free-of-charge multispectral images and their processing by applying limited manpower and resources. The testbed area is a strait between two Greek Islands in the Aegean Sea with many small islets and complex seabed relief. The HBT methodology implements semi-analytical and empirical steps to model sea-water inherent optical properties (IOPs) and apparent optical properties (AOPs) observed by the Sentinel-2A multispectral satellite. The relationships of the calculated IOPs and AOPs are investigated and utilized to classify the study area into sub-regions with similar water optical characteristics, where no environmental observations have previously been collected. The bathymetry model is configured using very few field data (training depths) chosen from existing official nautical charts. The assessment of the HBT indicates the potential for obtaining satellite derived bathymetry with a satisfactory accuracy for depths down to 30 m.