Initial Patch Research Articles

Digital surface model generation from high‐resolution satellite stereos based on hybrid feature fusion network

AbstractRecent studies have demonstrated that deep learning‐based stereo matching methods (DLSMs) can far exceed conventional ones on most benchmark datasets by both improving visual performance and decreasing the mismatching rate. However, applying DLSMs on high‐resolution satellite stereos with broad image coverage and wide terrain variety is still challenging. First, the broad coverage of satellite stereos brings a wide disparity range, while DLSMs are limited to a narrow disparity range in most cases, resulting in incorrect disparity estimation in areas with contradictory disparity ranges. Second, high‐resolution satellite stereos always comprise various terrain types, which is more complicated than carefully prepared datasets. Thus, the performance of DLSMs on satellite stereos is unstable, especially for intractable regions such as texture‐less and occluded regions. Third, generating DSMs requires occlusion‐aware disparity maps, while traditional occlusion detection methods are not always applicable for DLSMs with continuous disparity. To tackle these problems, this paper proposes a novel DLSM‐based DSM generation workflow. The workflow comprises three steps: pre‐processing, disparity estimation and post‐processing. The pre‐processing step introduces low‐resolution terrain to shift unmatched disparity ranges into a fixed scope and crops satellite stereos to regular patches. The disparity estimation step proposes a hybrid feature fusion network (HF2Net) to improve the matching performance. In detail, HF2Net designs a cross‐scale feature extractor (CSF) and a multi‐scale cost filter. The feature extractor differentiates structural‐context features in complex scenes and thus enhances HF2Net's robustness to satellite stereos, especially on intractable regions. The cost filter filters out most matching errors to ensure accurate disparity estimation. The post‐processing step generates initial DSM patches with estimated disparity maps and then refines them for the final large‐scale DSMs. Primary experiments on the public US3D dataset showed better accuracy than state‐of‐the‐art methods, indicating HF2Net's superiority. We then created a self‐made Gaofen‐7 dataset to train HF2Net and conducted DSM generation experiments on two Gaofen‐7 stereos to further demonstrate the effectiveness and practical capability of the proposed workflow.

On the Creation, Depletion, and End of Life of Polar Cap Patches

AbstractIonospheric convection patterns from the Super Dual Auroral Radar Network are used to determine the trajectories, transit times, and decay rates of three polar cap patches from their creation in the dayside polar cap ionosphere to their end of life on the nightside. The first two polar cap patches were created within 12 min of each other and traveled through the dayside convection throat, before entering the nightside auroral oval after 104 and 92 min, respectively. When the patches approached the nightside auroral oval, an intensification in the poleward auroral boundary occurred close to their exit point, followed by a decrease in the transit velocity. The last patch (patch 3) decayed completely within the polar cap and had a lifetime of only 78 min. After a change in drift direction, patch 3 had a radar backscatter power half‐life of 4.23 min, which reduced to 1.80 min after a stagnation, indicating a variable decay rate. 28 minutes after the change in direction, and 16 min after coming to a halt within the Clyde River radar field‐of‐view, patch 3 appeared to reach its end of life. We relate this rapid decay to increased frictional heating, which speeds up the recombination rate. Therefore, we suggest that the slowed patch motion within the polar cap convection pattern is a major factor in determining whether the patch survives as a recognizable density enhancement by the time the flux tubes comprising the initial patch cross into the nightside auroral oval.

A combustion mechanism reduction method based on entropy production analysis in fuel auto-ignition and laminar flames

This study provides a chemical mechanism reduction strategy based on entropy production analyses in both auto-ignition and laminar flames, which enhances the predictive accuracy for laminar burning velocities. In addition to chemical reactions, other irreversible sources causing entropy generation, such as mass diffusion and heat conduction, are considered in the modified approach. Specifically, initial skeletal mechanisms are first generated based on important reactions that contribute to entropy production in auto-ignition processes. Mechanism patches are then constructed to include important species and reactions, which contribute to entropy production from mass diffusion and heat conduction in laminar premixed flames beyond the pre-defined thresholds, respectively. Finally, the initial skeletal mechanisms and mechanism patches are combined to establish the final skeletal mechanisms. In this way, two final skeletal mechanisms for n-dodecane, consisting of 162 species and 2276 reactions, and 160 species and 1916 reactions, respectively, are developed from the detailed POLIMI mechanism with 451 species and 17,848 reactions. The two final skeletal mechanisms are proven to accurately predict laminar burning velocities and entropy production in n-dodecane flames with insignificant variations in the simulation results compared to the detailed mechanism, while their accuracy in predicting ignition delay times relies on the initial skeletal mechanisms. Specifically, the reduced mechanism with 160 species and 1916 reactions exhibits less satisfactory performance in predicting ignition delay compared to that with 162 species and 2276 reactions, indicating that a lower threshold is required to generate the initial skeletal mechanism through entropy production analysis of homogeneous auto-ignition processes. Additionally, compared with the reduced mechanisms with similar sizes obtained with other mechanism reduction strategies, the two final skeletal mechanisms accurately capture the characteristics of laminar burning velocities and ignition delay times, with similar calculation time required.

In situ observation and understanding of the arc erosion behavior of electrical contact materials

Abstract Arc erosion studies are carried out on electrical contact materials in the switching current experiment. A test rig is developed to capture the 2-D image and the 3-D profile of eroded silver tin oxide electrical contact material in situ. Image processing algorithms are specially presented to accurately recognize the features of erosion region. The initial small patch of erosion traces gradually evolves into the deep crater with the increment of switching cycle enabling a study of the arc erosion behavior. 2-D image, 3-D profile, and micro surface morphologies of the eroded contacts are analyzed and then related to the extent of erosion. The results show that the evolution process of central erosion region could be divided into the enlargement in breadth of contact surface and the extension in depth of contact profile. The end result of arc erosion is the net loss of contact material, which would lead to the substantial reduction of contact force and potential electrical contact instability.

Effectiveness of Discrete Piezo-patch Locations in Clamped-simply Supported Plates

Present study investigates the problem of determining the control effectiveness of piezoelectric patches for controlling vibration in higher modes of clamped-simply supported plates. A modal analysis approach is used to identify high average strain (or curvature) locations, as desired patch locations and represented in the form of geometrical parameters. The 9-patch configuration is used and modal analysis of integrated structure is carried out, which clearly shows that the initial patch configuration yields better effectiveness for square plates. In case of rectangular plates, the above strategy provides mixed results for the patch distribution effectiveness, for the modes considered here. Next, a concept of ‘stretching’ the patch distribution is applied for improving the effectiveness of the patch distribution and a Fig. of Merit (FOM), based on modal strains, is defined for quantifying the effectiveness. The results show that particular patch distribution geometry is suitable only for a specific group of modes. In view of this, a combined Fig. of Merit, based on the concept of modal weights, is also introduced, which has the potential to help in determining a suitable patch distribution through a formal optimization methodology. The control effectiveness comparison study of the initial and stretched patch configuration of plates also supports the quantification of control effectiveness of patch configuration through the FOM.

A weakly supervised inpainting-based learning method for lung CT image segmentation

Recently, various fully supervised learning methods are successfully applied for lung CT image segmentation. However, pixel-wise annotations are extremely expert-demanding and labor-intensive, but the performance of unsupervised learning methods are failed to meet the demands of practical applications. To achieve a reasonable trade-off between the performance and label dependency, a novel weakly supervised inpainting-based learning method is introduced, in which only bounding box labels are required for accurate segmentation. Specifically, lesion regions are first detected by an object detection network, then we crop them out of the input image and recover the missing holes to normal regions using a progressive CT inpainting network (PCIN). Finally, a post-processing method is designed to get the accurate segmentation mask from the difference image of input and recovered images. In addition, real information (i.e., number, location and size) of the bounding boxes of lesions from the dataset guides us to make the training dataset for PCIN. We apply a multi-scale supervised strategy to train PCIN for a progressive and stable inpainting. Moreover, to remove the visual artifacts resulted from the invalid features of missing holes, an initial patch generation network (IPGN) is proposed for holes initialization with generated pseudo healthy image patches. Experiments on the public COVID-19 dataset demonstrate that PCIN is outstanding in lung CT images inpainting, and the performance of our proposed weakly supervised method is comparable to fully supervised methods.

D-UAP: Initially Diversified Universal Adversarial Patch Generation Method

With the rapid development of adversarial example technologies, the concept of adversarial patches has been proposed, which can successfully transfer adversarial attacks to the real world and fool intelligent object detection systems. However, the real-world environment is complex and changeable, and the adversarial patch attack technology is susceptible to real-world factors, resulting in a decrease in the success rate of attack. Existing adversarial-patch-generation algorithms have a single direction of patch initialization and do not fully consider the impact of initial diversification on its upper limit of adversarial patch attack. Therefore, this paper proposes an initial diversified adversarial patch generation technology to improve the effect of adversarial patch attacks on the underlying algorithms in the real world. The method uses YOLOv4 as the attack model, and the experimental results show that the attack effect of the adversarial-patch-attack method proposed in this paper is higher than the baseline 8.46%, and it also has a stronger attack effect and fewer training rounds.

Modeling and Simulation of a Micro strip Antenna in Annular Geometry

Micro strip antenna has been widely developed over time due to its flexibility and easier todesign. The aim of this work is the simulation of an Annular Micro strip Patch antenna using CST software(Computer Simulation Technology) in order to know its performance (Gain, the reflection parameter S11,directivity, efficiency…). Adding slots to the initial patch was determined to improve the main factors ofthese characteristics.

Illposedness of $$C^{2}$$ Vortex Patches

It is well known that vortex patches are wellposed in $C^{1,\alpha}$ if $0<\alpha <1$. In this paper, we prove the illposedness of $C^{2}$ vortex patches. The setup is to consider the vortex patches in Sobolev spaces $W^{2,p}$ where the curvature of the boundary is $L^p$ integrable. In this setting, we show the persistence of $W^{2,p}$ regularity when $1<p <\infty$ and construct $C^{2}$ initial patch data for which the curvature of the patch boundary becomes unbounded immediately for $t>0$. The key ingredient is the evolution equation for the curvature, the dominant term in which turns out to be linear and dispersive.

Vanishing meadows — Quantitative analysis of factors driving population declines of endangered butterflies

There is a growing number of alarming reports indicating serious declines in biodiversity of invertebrates, mostly manifested through loss of local populations and abundances, and prognosing their regional extinctions. Proper defining of threats is crucial for the effective conservation of endangered species and valuable habitats. Nevertheless, while typically such threats are broadly recognized, their long-term quantitative assessments are rarely available. The aim of this study was to identify and quantify the main drivers of population dynamics and habitat loss of two endangered butterfly species of the genus Phengaris within an extensive semi-urban landscape with 70 initial habitat patches. Over two decades of the study, almost half of the habitat patches remained abandoned, whereas a quarter was directly destroyed through conversion to build-up areas. The latter process was most harmful to butterfly numbers, although it occurred only outside Natura 2000 protected sites. Besides, afforestation works, even if rare, caused significant declines in local populations of the investigated butterflies. Similarly, the meadow abandonment led to decrease in butterfly numbers and their local extinctions due to habitat loss. Meadows were threatened mostly by direct human destruction. The habitat area decreased nearly by 15 %, of which ca. 2/3 suffered direct destruction by built-up, especially in most recent years. As meadows are one of globally threatened terrestrial biodiversity hotspots, and meadow abandonment is partially reversible, there is an urgent need to implement conservation-oriented management of remaining fragments as an effective tool ensuring the persistence of these habitats and their species communities.

How dirt cones form on glaciers: Field observation, laboratory experiments, and modeling.

Dirt cones are meter-scale structures encountered at the surface of glaciers, which consist of ice cones covered by a thin layer of ashes, sand, or gravel, and which form naturally from an initial patch of debris. In this article, we report field observations of cone formation in the French Alps, laboratory-scale experiments reproducing these structures in a controlled environment, and two-dimensional discrete-element-method-finite-element-method numerical simulations coupling the grain mechanics and thermal effects. We show that cone formation originates from the insulating properties of the granular layer, which reduces ice melting underneath as compared to bare ice melting. This differential ablation deforms the ice surface and induces a quasistatic flow of grains that leads to a conic shape, as the thermal length become small compared to the structure size. The cone grows until it reaches a steady state in which the insulation provided by the dirt layer exactly compensates for the heat flux coming from the increased external surface of the structure. These results allowed us to identify the key physical mechanisms at play and to develop a model able to quantitatively reproduce the various field observations and experimental findings.

New quantum algorithm for visual tracking

Visual tracking, which trains a classifier to distinguish the target from the surrounding environment given an initial sample patch containing the target, plays an important role in computer vision. Yu et al. proposed a quantum algorithm for visual tracking (QVT) [Phys. Rev. A 94, 042311 (2016)] with time complexity OϱZϱZ+ϱX2polylogNϵ based on the framework proposed by Henriques et al. [IEEE Trans. Pattern Anal. Mach. Intell. 7, 583 (2015)], where ϱXZ is the condition number of the data matrix XZ, N is the dimension of the original sample patch, and ϵ is the desired accuracy of the output state. To get a further speedup, we propose a new QVT with time complexity OϱZ1+ϱXpolylogNϵ based on the algorithm of Henriques et al. Our algorithm achieves a quadratic speedup on the condition number ϱX(Z) compared to the algorithm of Yu et al. Also, it shows exponential speedups on N over the classical counterpart when ϱX(Z) and ϵ are OpolylogN. Finally, we extend it to the nonlinear two-dimensional multi-channel case.

Modelling vegetation land fragmentation in urban areas of Western Province, Sri Lanka using an Artificial Intelligence-based simulation technique.

Vegetation land fragmentation has had numerous negative repercussions on sustainable development around the world. Urban planners are currently avidly investigating vegetation land fragmentation due to its effects on sustainable development. The literature has identified a research gap in the development of Artificial Intelligence [AI]-based models to simulate vegetation land fragmentation in urban contexts with multiple affecting elements. As a result, the primary aim of this research is to create an AI-based simulation framework to simulate vegetation land fragmentation in metropolitan settings. The main objective is to use non-linear analysis to identify the factors that contribute to vegetation land fragmentation. The proposed methodology is applied for Western Province, Sri Lanka. Accessibility growth, initial vegetation large patch size, initial vegetation land fragmentation, initial built-up land fragmentation, initial vegetation shape irregularity, initial vegetation circularity, initial building density, and initial vegetation patch association are the main variables used to frame the model among the 20 variables related to patches, corridors, matrix and other. This study created a feed-forward Artificial Neural Network [ANN] using R statistical software to analyze non-linear interactions and their magnitudes. The study likewise utilized WEKA software to create a Decision Tree [DT] modeling framework to explain the effect of variables. According to the ANN olden algorithm, accessibility growth has the maximum importance level [44] between -50 and 50, while DT reveals accessibility growth as the root of the Level of Vegetation Land Fragmentation [LVLF]. Small, irregular, and dispersed vegetation patches are especially vulnerable to fragmentation. As a result, study contributes detech and managing vegetation land fragmentation patterns in urban environments, while opening up vegetation land fragmentation research topics to AI applications.

A distance-driven urban simulation model (DISUSIM): Accounting for urban morphology at multiple landscape levels

Data and skill gaps are common in exploring future urban growth scenarios, especially in developing countries. This study demonstrates how to use distance-related laws regarding process and pattern of urban development to bypass the heavy data and skills required in traditional urban expansion models that rely on various driving factors and machine-learning algorithms. We proposed an urban simulation model that uses three distance-driven components alone to simulate urban expansion by regulating urban morphology at multiple levels. The landscape-level component discriminates the spatiotemporal distribution of urban demand based on rules linked to distance to city centers. The class-level component applies an exponential model with the distance to pre-urbanized patches to control where new urban development may expand. The patch-level component creates urban patches of given sizes with shapes controlled by their distances to the initial patch seeds. Application of the model in Wuhan, China confirms the efficacy of the model and its distance-driven components, as supported by both cell-level agreement and pattern-level similarity. The overarching contribution of DisUSIM lies in providing a distance-driven framework for convenient exploration of urban dynamics in situations where datasets and skills on processing and analyzing driving factors and mechanisms of urban expansion are not readily available.

An extended and improved particle‐spectral method for analysis of unsteady inviscid incompressible flows through a channel of finite length

AbstractThis paper addresses the development of a method to simulate and analyze unsteady flows of incompressible and inviscid fluids in rectangular channels. Mathematically, the flow problem formulates as Euler equations in terms of the stream function and vorticity. The normal velocity and the vorticity are given at the inlet of the channel, and only the normal velocity is specified at the outlet. The particle‐spectral method describes the vorticity field change in a Lagrangian sense by dynamics of marker particles and an Eulerian description of the velocity field using the global Galerkin approximation. The method involves the subdivision of flow domain on rectangular cells; the approximation of vorticity distribution in cells by least‐squares method; Galerkin coefficients calculation using analytical integration on space variables. For time integration, the five‐stage Runge‐Kutta method of the third accuracy order and the sixth one of symplecticity with adaptive step size control uses. The extension of the particle‐spectral approach comprises algorithms for fluid flows qualitative analysis jointly with the time simulation of unsteady flows. This includes the trajectories of marker particles calculation, computing the particle's residence time in the channel, kinetic energy dynamics, and algorithms for analyzing the dynamics and mixing of vortex patches. Test calculations based on reference solutions showed the effectiveness and good accuracy of the particle‐spectral approach. Using the developed method the dynamics of one and two initial vortex patches in the channel were investigated. We found the final state of their long‐time dynamics is time‐periodic or quasi‐periodic movements. The application of qualitative analysis algorithms demonstrates these final states produce a chaotic scattering of fluid particles in the flow‐through zone of the channel.

The evolution of the code during review: an investigation on review changes

Code review is a software engineering practice in which reviewers manually inspect the code written by a fellow developer and propose any change that is deemed necessary or useful. The main goal of code review is to improve the quality of the code under review. Despite the widespread use of code review, only a few studies focused on the investigation of its outcomes, for example, investigating the code changes that happen to the code under review. The goal of this paper is to expand our knowledge on the outcome of code review while re-evaluating results from previous work. To this aim, we analyze changes that happened during the review process, which we define as review changes. Considering three popular open-source software projects, we investigate the types of review changes (based on existing taxonomies) and what triggers them; also, we study which code factors in a code review are most related to the number of review changes. Our results show that the majority of changes relate to evolvability concerns, with a strong prevalence of documentation and structure changes at type-level. Furthermore, differently from past work, we found that the majority of review changes are not triggered by reviewers’ comments. Finally, we find that the number of review changes in a code review is related to the size of the initial patch as well as the new lines of code that it adds. However, other factors, such as lines deleted or the author of the review patchset, do not always show an empirically supported relationship with the number of changes.

Numerical simulation of vegetation evolution in compound channels.

The study of vegetation evolution is essential for further understanding of biogeographic feedback and ecological restoration. In this paper, a vegetation evolution model based on velocity threshold (scaled by the average flow velocity of the bare bed) was established to simulate the vegetation evolution process in compound channels. In this model, the effect of vegetation on water flow was generalized as equivalent Manning coefficient, and the velocity field was obtained by solving two-dimensional shallow water equations. The model defined that new vegetation was added in areas where the velocity was less than the velocity threshold, and conversely, vegetation was destroyed in areas where the flow velocity exceeded the velocity threshold. The model was used to explore the effect of velocity threshold, initial vegetation coverage, and relative water depth (the ratio of the flow depth in the floodplain to that over main channel) on final vegetation coverage and longitudinal dispersion coefficient (Ke) in compound channels, and compare the difference of vegetation evolution between rectangular channels and compound channels. Results showed that the velocity threshold played a decisive role in vegetation evolution, and the effect of relative water depth and cross section type on vegetation evolution was only reflected when the velocity threshold was small. The longitudinal dispersion coefficient gradually increased with the expansion of vegetation, and tended to a constant value (Kf) when a stable vegetation landscape was reached. As the relative water depth decreased, the longitudinal dispersion coefficient presented an increasing trend. Regular distribution of initial vegetation patches can produce larger longitudinal dispersion coefficient compared to the cases of random distribution in compound and rectangular channels, and the increasing effect was more significant in compound channels.

2x2 Array Microstrip Patch Antenna Designing and Validation with Artificial Neural Networks

A novel and successful method for RF, microwave modelling, and antenna design has gained attention: computational modules based on neural networks. The behavior of active or passive components or circuits can be taught to neural networks. The design of a microstrip patch antenna utilizing an artificial neural network (ANN) is discussed in this research. Computer Simulation Technology (CST) is used to establish the antenna's size and characteristics in order to create a successful ANN model. The feed-forward back-propagation neural network and Levenberg Marquardt optimization algorithm are used to model the antenna design. High-Frequency Structure Simulator (HFSS) software, operating at 2.4 GHz, is used to build the initial microstrip patch (ISM band).A 2.4GHz operating frequency is used to create a 2*2 microstrip planar array. A variety of neural networks are evaluated and trained to produce the most ideal results after generating and modelling discoveries using HFSS simulation software and the Finite Difference Time Domain (FTDT) approach. In this research, neural networks are utilized. In order to identify the most ideal solution, optimization is performed using a radial basis function neural network (RBF NN) and a feed-forward backpropagation approach.

Can Cervical and Lumbar Epidural Blood Patches Help Avoid Revision Surgery for Symptomatic Postoperative Dural Tears?

Incidental durotomies resulting in symptomatic postoperative cerebrospinal fluid (CSF) leaks can be treated with bedrest, subarachnoid lumbar drain, and surgical re-exploration. Another option is an epidural blood patch, which forms a clot over the dural tear in a minimally invasive manner. Our objective was to describe our center's outcomes and complications following epidural blood patches for symptomatic postoperative durotomies. Patients undergoing spine surgeries at one institution from 2012-2020 were included. Patient charts were reviewed for demographic information, surgical data, rate of dural tear, type of repair, and use of blood patches postoperatively. A total of 1392 patients (726 male, 666 female, average age: 56.3 ± 15.4 years), including 436 cervical and 956 posterior thoracolumbar/lumbar spine surgeries were screened. There were 6 (1.4%) cervical, 64 (6.7%) lumbar, and a total of 70 (5.0%) patients with incidental dural tears/blebs. Of these patients, we identified 2 cervical and 8 lumbar spine patients with persistent CSF leaks who received epidural blood patches postoperatively. Nine of 10 (90%) had resolution of symptoms. One lumbar patient failed both an initial blood patch and subsequent surgical re-exploration, but had successful relief after a second blood patch. The incidence of dural tears/blebs was 1.4% in cervical, 6.7% in lumbar, and 5.0% in all spine surgeries. Of the 10 patients with symptomatic CSF leaks, 9 were successfully treated with blood patches. Targeted epidural blood patch is effective in treating symptomatic CSF leaks and minimizes the morbidity of surgical re-exploration.

Algorithmic Multimodality Approach to Diagnosis and Treatment of Spinal CSF Leak and Venous Fistula in Patients With Spontaneous Intracranial Hypotension.

Spontaneous intracranial hypotension (SIH) is a disorder of CSF dynamics that causes a complex clinical syndrome and severe disability. SIH is challenging to diagnose because of the variability of its presenting clinical symptoms, the potential for subtle imaging findings to be easily overlooked, and the need for specialized diagnostic testing. Once SIH is suggested by clinical history and/or supported by initial neuroim-aging, many patients may undergo initial nontargeted epidural blood patching with variable and indefinite benefit. However, data suggest that precise localization of the CSF leak or CSF-venous fistula (CVF) can lead to more effective and durable treatment strategies. Leak localization can be achieved using a variety of advanced diagnostic imaging techniques, although these may not be widely performed at nontertiary medical centers, leaving many patients with the potential for inadequate workup or treatment. This review describes imaging techniques including dynamic fluoroscopic and CT myelography as well as delayed MR myelography and treatment options including percutaneous, endovascular, and surgical approaches for SIH. These are summarized by an algorithmic framework for radiologists to approach the workup and treatment of patients with suspected SIH. The importance of a multidisciplinary approach is emphasized.