Refined UNet V4: End-to-End Patch-Wise Network for Cloud and Shadow Segmentation with Bilateral Grid

Remote sensing images are usually contaminated by cloud and corresponding shadow regions, making cloud and shadow detection one of the essential prerequisites for processing and translation of remote sensing images. Edge-precise cloud and shadow segmentation remains challenging due to the inherent high-level semantic acquisition of current neural segmentation fashions. We, therefore, introduce the Refined UNet series to partially achieve edge-precise cloud and shadow detection, including two-stage Refined UNet, v2 with a potentially efficient gray-scale guided Gaussian filter-based CRF, and v3 with an efficient multi-channel guided Gaussian filter-based CRF. However, it is visually demonstrated that the locally linear kernel used in v2 and v3 is not sufficiently sensitive to potential edges in comparison with Refined UNet. Accordingly, we turn back to the investigation of an end-to-end UNet-CRF architecture with a Gaussian-form bilateral kernel and its relatively efficient approximation. In this paper, we present Refined UNet v4, an end-to-end edge-precise segmentation network for cloud and shadow detection, which is capable of retrieving regions of interest with relatively tight edges and potential shadow regions with ambiguous edges. Specifically, we inherit the UNet-CRF architecture exploited in the Refined UNet series, which concatenates a UNet backbone of coarsely locating cloud and shadow regions and an embedded CRF layer of refining edges. In particular, the bilateral grid-based approximation to the Gaussian-form bilateral kernel is applied to the bilateral message-passing step, in order to ensure the delineation of sufficiently tight edges and the retrieval of shadow regions with ambiguous edges. Our TensorFlow implementation of the bilateral approximation is relatively computationally efficient in comparison with Refined UNet, attributed to the straightforward GPU acceleration. Extensive experiments on Landsat 8 OLI dataset illustrate that our v4 can achieve edge-precise cloud and shadow segmentation and improve the retrieval of shadow regions, and also confirm its computational efficiency.

PARTICLE TRAJECTORIES AND THE PERCEPTION OF CLASSICAL MOTION IN THE FREE PROPAGATION OF WAVE PACKETS.

The free propagation in time of a normalisable wave packet is the oldest problem of continuum quantum mechanics. Its motion from microscopic to macroscopic distance is the way in which most quantum systems are detected experimentally. Although much studied and analysed since 1927 and presented in many text books, here the problem is re-appraised from the standpoint of semi-classical mechanics. Particular aspects are the emergence of deterministic trajectories of particles emanating from a region of atomic dimensions and the interpretation of the wave function as describing a single particle or an ensemble of identical particles. Of possible wave packets, that of gaussian form is most studied due to the simple exact form of the time-dependent solution in real and in momentum space. Furthermore, this form is important in laser optics. Here the equivalence of the time-dependent Schroedinger equation to the paraxial equation for the propagation of light is demonstrated explicitly. This parallel helps to understand the relevance of trajectory concepts and the conditions necessary for the perception of motion as classical.

Analysis of the Spatial-Frequency Characteristics of the Photo-Assisted Method of a Quartz Rough Surface Nano-Polishing

The relationship between the spatial-frequency parameters of a rough surface with a random profile, which has a Gaussian form of the correlation function, and the amplitude-frequency characteristics of the electric field created by this surface is determined. The numerical determination of the evanescent field optimal configuration formed near the quartz rough surface in the gaseous medium saturated with chlorine molecules when illuminated from the quartz side has been considered. The finite-element approach is used to solve the Helmholtz two-dimensional vector equation. It was found that at the initial stage of photochemical polishing different electrodynamic conditions are created for the etching process depending on the profile height standard deviation value. In particular, when the standard deviation is less than 1 nm, all surface protrusions, for which the spatial spectrum harmonics of the profile are located in the region of the maximum slope of the spectral function, are most actively etched. This leads to a decrease in the effective width of the spatial spectrum of a rough quartz surface and an increase in its correlation length. Therefore, simultaneously with decreasing the height of the protrusions, the surface becomes flatter. The paper shows the different character of quartz surface nano-polishing process conditions depending on the initial standard deviation of the profile height.

Superstatistical modelling of protein diffusion dynamics in bacteria

A recent experiment (Sadoon AA, Wang Y. 2018 Phys. Rev. E 98 , 042411. ( doi:10.1103/PhysRevE.98.042411 )) has revealed that nucleoid-associated proteins (i.e. DNA-binding proteins) exhibit highly heterogeneous diffusion processes in bacteria where not only the diffusion constant but also the anomalous diffusion exponent fluctuates for the various proteins. The distribution of displacements of such proteins is observed to take a q -Gaussian form, which decays as a power law. Here, a statistical model is developed for the diffusive motion of the proteins within the bacterium, based on a superstatistics with two variables. This model hierarchically takes into account the joint fluctuations of both the anomalous diffusion exponents and the diffusion constants. A fractional Brownian motion is discussed as a possible local model. Good agreement with the experimental data is obtained.

Statistical treatment of signal power measurements under identification of the distant radiator antenna directional pattern

Based on the use of the least squares method and the fastest gradient descent algorithm, the procedure for approximation of measurements of the radiation power of the transmitting antenna has been developed. This procedure involves an average power pattern of a linear type antenna with random amplitude and phase deviations on the assumptions of their normal distribution, consistency of average amplitude and phase dispersion along the antenna, Gaussian form of the correlation coefficient of deviations. As an approximation quality criterion, the minimum of the sum of the squared deviations of radiation power estimates at points of measurement from theoretical values of a random function that determines the shape of the radiation pattern is used. Given the nonlinearity of the function, the numerical approximation procedure based on the fastest gradient descent method has been proposed. The direction of descent at the next iteration is selected on the basis of the current value of the gradient of the criterion function. The criterion for the completion of the process is the achievement of a stationary, in a statistical sense, area of the function.

Limitation of the frequency dispersion of electromagnetic radiation produced by fluxons.

The work found the shape of spectral lines of electromagnetic radiation produced by a relativistic quantum of magnetic flux moving in a Josephson transmission line under the action of a purely fluctuating external current and dissipation. It is shown that even in the case of a Gaussian noise current, the shape of the spectral lines of electromagnetic radiation has an essentially non-Gaussian form. The limited dispersion of the frequency of electromagnetic radiation (which is determined by the pinching of the spectral lines of radiation) is proved, due to the relativistic properties of the quantum magnetic flux. This is another unique aspect of the relativistic quantum magnetic flux - as a carrier of electromagnetic radiation in the Josephson transmission line.

Millimeter-wave Techniques: Theory, Algorithms and Methods

The goal of book is to present the theory, algorithms and methods to design the millimeter-wave Gyrotron beam-shaping antenna system (of good metal conductor, e.g., of aluminum or copper) to shape four roughly Gaussian beams of different frequencies into one specified high-quality Gaussian beam at the output of the mirror system. The input beams are from a quasi-optical launcher which converts each of the following mode/frequency pairs into a roughly Gaussian form: TE21,6 (107.5 GHz), TE22,6 (110.0 GHz), TE24,7 (124.5 GHz) and TE25,7 (127.5 GHz). Each of the output beams from the launcher is to be converted into a Gaussian beam with specified waist radius and with as high a coupling coefficient as possible to an ideal Gaussian beam. A design optimization procedure for the mirror system is developed and applied to the case of the four mode/frequency pairs above. The frequencies are specified because they are of interest to General Atomics. The 127.5 GHz frequency is the ITER start-up frequency. In addition, the design obtained is tested by simulation for two other mode/frequency pairs, TE23,6 (112.9 GHz) and TE23,7 (121.5 GHz). The theory, algorithms and methods required in such multi-mode beam-shaping mirror system designs are investigated. These include (1) the iterative phase retrieval technique for phase retrieval from measured magnitude-only data; (2) a theoretical formula developed to evaluate the validity of the image theorem approximation in the cylindrical geometry used during the mirror system design; (3) the Taylor-FFT algorithm developed in this research work for fast computation of the electromagnetic wave propagation and scattering; (4) the newly developed phase gradient phase correction method for mirror surface correction; and (5) the least mean square optimization method, proposed for multi-mode mirror design. Based on our theory, algorithms and methods, a coupling coefficient to the target Gaussian beam greater than 99.99% is achieved for the single-mode designs and an average coupling coefficient up to 99.50% within the window aperture of 88 mm diameter for the multi-mode design. The design has been successfully verified the Method of Moments software Surf3D.

A Novel Robust Student’s t-Based Cubature Information Filter with Heavy-Tailed Noises

In this paper, a novel robust Student’s t-based cubature information filter is proposed for a nonlinear multisensor system with heavy-tailed process and measurement noises. At first, the predictive probability density function (PDF) and the likelihood PDF are approximated as two different Student’s t distributions. To avoid the process uncertainty induced by the heavy-tailed process noise, the scale matrix of the predictive PDF is modeled as an inverse Wishart distribution and estimated dynamically. Then, the predictive PDF and the likelihood PDF are transformed into a hierarchical Gaussian form to obtain the approximate solution of posterior PDF. Based on the variational Bayesian approximation method, the posterior PDF is approximated iteratively by minimizing the Kullback-Leibler divergence function. Based on the posterior PDF of the auxiliary parameters, the predicted covariance and measurement noise covariance are modified. And then the information matrix and information state are updated by summing the local information contributions, which are computed based on the modified covariance. Finally, the state, scale matrix, and posterior densities are estimated after fixed point iterations. And the simulation results for a target tracking example demonstrate the superiority of the proposed filter.

Power Spectrum in the Conductive Terrestrial Ionosphere

Stochastic differential equation of the phase fluctuations is derived for the collision conductive magnetized plasma in the polar ionosphere applying the complex geometrical optics approximation. Calculating second order statistical moments it was shown that the contribution of the longitudinal conductivity substantially exceeds both Pedersen and Hall’s conductivities. Experimentally observing the broadening of the spatial power spectrum of scattered electromagnetic waves which equivalent to the brightness is analyzed for the elongated ionospheric irregularities. It was shown that the broadening of the spectrum and shift of its maximum in the plane of the location of an external magnetic field (main plane) less than in perpendicular plane for plasmonic structures having linear scale tenth of kilometer; and substantially depends on the penetration angle of an incident wave in the conductive collision turbulent magnetized ionospheric plasma. The angle-of-arrival (AOA) in the main plane has the asymmetric Gaussian form while in the perpendicular plane increases at small anisotropy factors and then tends to the saturation for the power-low spectrum characterizing electron density fluctuations. Longitudinal conductivity fluctuations increase the AOAs of scattered radiation than in magnetized plasma with permittivity fluctuations. Broadening of the temporal spectrum containing the drift velocity of elongated ionospheric irregularities in the polar ionosphere allows to solve the reverse problem restoring experimentally measured velocity of the plasma streams and characteristic linear scales of anisotropic irregularities in the terrestrial ionosphere.

Asymptotic Analysis of Retrial Queueing System M/M/1 with Impatient Customers, Collisions and Unreliable Server

The retrial queueing system of M=M=1 type with Poisson flow of arrivals, impatient cus- tomers, collisions and unreliable service device is considered in the paper. The novelty of our contribution is the inclusion of breakdowns and repairs of the service into our previous study to make the problem more realistic and hence more complicated. Retrial time of customers in the orbit, service time, impa- tience time of customers in the orbit, server lifetime (depending on whether it is idle or busy) and server recovery time are supposed to be exponentially distributed. An asymptotic analysis method is used to find the stationary distribution of the number of customers in the orbit. The heavy load of the system and long time patience of customers in the orbit are proposed as asymptotic conditions. Theorem about the Gaussian form of the asymptotic probability distribution of the number of customers in the orbit is formulated and proved. Numerical examples are given to show the accuracy and the area of feasibility of the proposed method