Multi-scale Frame Research Articles

Object Detection in Video Surveillance Based on Multiscale Frame Representation and Block Processing by a Convolutional Neural Network

Object Detection in Video Surveillance Based on Multiscale Frame Representation and Block Processing by a Convolutional Neural Network

Prediction of atmospheric motion vectors around typhoons using generative adversarial network

In this study, atmospheric motion vectors (AMVs) were derived from the satellite images predicted using a generative adversarial network (GAN) and a deep multi-scale frame prediction algorithm. The GAN was trained and tested with a sequence of the satellite images of a COMS satellite infrared-window channel under the 68 tropical cyclones. The inputs of the consecutive satellite images with 15-min interval were then processed using the trained GAN model to generate satellite images in the next time steps. To further enhance the model’s predictability, particle image velocimetry based on the theory of cross-correlation schemes was employed to the GAN-generated satellite image sequence and AMVs were produced. The GAN-derived AMVs were validated with the wind fields based on the numerical weather prediction (NWP) and radiosonde observations. The comparisons showed that the GAN-derived AMVs depicted the structure of atmospheric circulations with a certain level of accuracy. Through comparison with the radiosonde observations, the root-mean-square error and the wind speed bias of the GAN-derived AMVs were comparable to, and even smaller than those of the NWP-derived wind fields. The current approach may enhance the accuracy in predicting short-term wind velocity fields, which in turn may provide more realistic inputs in storm surge modeling.

Developed approach for phase-based Eulerian video magnification

This paper proposes a modification approach for phased-based EVM in order to reduce the processing time without effect the quality of the magnified video. The proposed approach applies a resizing process on the input video using Lanczos-3 algorithm. Then, it decomposes video frames using steerable pyramid to obtain multi-scale frame with its orientation. Subsequently, the resulted frames are filtered by temporal filters for specific bands and the filtered frames are multiplied by a magnification factor. Now, both the magnified regions and the unmagnified regions for each frame are added together. Finally, reconstructing the produced magnified multi-scale frames using the inverse steerable pyramid. The experimental results show that superiority of the proposed approach compares to the conventional phase-based EVM in processing time, where the processing time reduction about 60-65%. Furthermore, this approach does not affect on the video quality, which maintain it in the boundary of the conventional Phase-based EVM.

Adversarial learning for deformable registration of brain MR image using a multi-scale fully convolutional network

Background and objectiveDeformable registration is very significant for various clinical image applications. Unfortunately, existing conventional medical image registration approaches, which involve time-consuming iterative optimization, have not reached the level of routine clinical practice in terms of registration time and robustness. The aim of this study is to propose a tuning-free 3D image registration model based on adversarial deep network, and to achieve rapid and high-accurate registration. MethodsWe propose a fully convolutional network (FCN) to regress the 3D dense deformation field in one shot from the to-be-registered image pair. To precisely regress the complex deformation and produce optimal registration, we design the FCN as a novel multi-scale frame to capture the complementary multi-scale image features and effectively characterize the spatial correspondence between the image pair. Moreover, we learn a discriminator network simultaneously to discriminate the registered two images, where the discrimination loss helps further update the FCN. Thus by the adversarial training strategy, the registration network is urged to produce well-registered two images that are indistinguishable for the discriminator. ResultsWe perform registration experiments on four different brain MR datasets using the model trained by ANDI database. Compared with some state-of-the-art registration algorithms including other newest deeplearning-based methods, the proposed method provides a considerable increase of large than 4% in terms of Dice similarity coefficient (DSC). Moreover, our model also obtains comparable distance errors. More significantly, our model can achieve a high-accurate 3D registration result in average 0.74 s, with roughly hundred speed-up over conventional registration methods. ConclusionsThe proposed model shows consistent high performance for various registration tasks under a second without any additional parameter tuning, which proves its potential for real-time clinical applications.

Computational Life Assessment of ASR-damaged RC Decks by Site-Inspection Data Assimilation

To estimate the remaining life of existing RC bridge decks damaged by alkali silica reaction (ASR), multi-scale numeri-cal analysis with chemo-hygral model is integrated with visual inspection data at site. First, the applicability of the poro-mechanical models for ASR expansion in the multi-scale frame are examined with the experiments of the real scale RC slabs and the model is validated to bring about fair prediction of the 3D anisotropic expansion and the fatigue life of the slabs. Second, visually inspected cracks on bottom surfaces of RC decks are converted to space-averaged strains, and the magnitude of ASR is estimated from the vertical deformation, based on which the internal pre-stress and the damage fields are re-produced by numerical predictor-corrector cycles, and the remaining life of ASR damaged RC bridge decks is fairly estimated. By conducting sensitivity analyses in terms of ASR-gel volumes and cracks, allowable error range of site inspection data is clarified to meet the requirement of asset management.

A Multiscale Approach to the Smart Deployment of Micro-Sensors over Lightweight Structures.

A topology optimization approach has been recently proposed to maximize the sensitivity to damage of measurements, collected through a network of sensors to be deployed over thin plates for structural health monitoring purposes. Within such a frame, damage is meant as a change in the structural health characterized by a reduction of relevant stiffness and load-carrying properties. The sensitivity to a damage of unknown amplitude and location is computed by comparing the response to the external actions of the healthy structure and of a set of auxiliary damaged structures, each one featuring reduced mechanical properties in a small region only. The topology optimization scheme has been devised to properly account for the information coming from all of the sensors to be placed on the structure and for damage depending on its location. In this work, we extend the approach within a multiscale frame to account for three different length scales: a macroscopic one, linked to the dimensions of the whole structure to be monitored; a mesoscopic one, linked to the characteristic size of the damaged region; a microscopic one, linked to the size of inertial microelectromechanical systems (MEMS) to be used within a marginally-invasive health monitoring system. Results are provided for a square plate and for a section of fuselage with stiffeners, to show how the micro-sensors have to be deployed to maximize the capability to detect a damage, to assess the sensitivity of the results to the measurement noise and to also discuss the speedup in designing the network topology against a standard single-scale approach.

Microstructural characterization of creep anisotropy at 673 K in the M5® alloy

Zirconium alloy tubing is used in pressurized water nuclear reactors in order to prevent fissile material from leaking into the coolant. It can be the first safety wall of nuclear fuel, and is submitted to complex thermomechanical loadings. In consequence, new Nb-modified alloys, such as the M5® alloy, and fine numerical models are being developed to guarantee a better and longer mechanical integrity of these tubes. To identify the physical mechanisms that could be considered in such models, an experimental approach, combining creep tests with electron backscattered diffraction and Transmission electron microscopy investigations, was carried out.Tubular specimens were submitted to multiaxial creep tests at a temperature of 673K. Seven ratios between the axial and hoop applied stresses were investigated. It enabled a macroscopic evidence of the creep anisotropy. Besides, EBSD analyses on a mesoscopic-sized non deformed area led to the characterization of the variation of the average Schmid factor with the direction of loading. Finally, TEM observations were done on seven crept samples, corresponding to the seven directions of loading tested mechanically. The variations of the different parameters investigated (activated slip systems, dislocation densities, curvatures of the dislocations) can be seen as the effects of the creep anisotropy at a microscopic scale. The correlation between results is then discussed in a multiscale frame.

Structural Design for Dynamic Roller Device in Transporting Unit of Biomass Pyrolytic System

With the full study of industrial process control technology for thermal, chemical industry, the application process, the FCS system of biomass pyrolysis unit was development combined with the actual. Feeding system uses the weighing system of dynamic belt with an advanced embedded technology. The computing speed and accuracy, data processing compared with traditional electronic belt scale, information exchange, integration, reliability and stability will be greatly improved. The technology shall be constantly upgraded and vigorously promoted. The frame structures of dynamic roller were various in forms. Different systems can be used to the frame with different scale structures. The suspension of multi-scale frame of the roller was used widely currently. The modeling and structural optimization of the frames in reasonable scale were designed with reducing the use of metal materials, optimizing the overall performance of mechanical systems and improving the weighing accuracy. The analysis of dynamic roller in scale weighing and speed principle, the establishment of a multi-roller single floating frame structure of the scale model and the analysis of the roller of the force were built. The verification in final design through the ANSYS software is reasonable.

Multi-scale algorithm for improved scintillation detection in a CCD-based gamma camera

Gamma cameras based on charge-coupled devices (CCDs) and micro-columnar CsI scintillators can reach high spatial resolutions. However, the gamma interaction probability of these scintillators is low (typically <30% at 141 keV) due to the limited thickness of presently available micro-columnar scintillators. Continuous scintillators can improve the interaction probability but suffer from increased light spread compared to columnar scintillators. In addition, for both types of scintillators, gamma photons incident at an oblique angle reduce the spatial resolution due to the variable depth of interaction (DOI). To improve the spatial resolution and spectral characteristics of these detectors, we have developed a fast analytic scintillation detection algorithm that makes use of a depth-dependent light spread model and as a result is able to estimate the DOI in the scintillator. This algorithm, performing multi-scale frame analysis, was tested for an electron multiplying CCD (EM-CCD) optically coupled to CsI(Tl) scintillators of different thicknesses. For the thickest scintillator (2.6 mm) a spatial resolution of 148 µm full width half maximum (FWHM) was obtained with an energy resolution of 46% FWHM for perpendicularly incident gamma photons (interaction probability 61% at 141 keV). The multi-scale algorithm improves the spatial resolution up to 11%, the energy resolution up to 36% and the signal-to-background counts ratio up to 46% compared to a previously implemented algorithm that did not model the depth-dependent light spread. In addition, the multi-scale algorithm can accurately estimate DOI. As a result, degradation of the spatial resolution due to the variable DOI for gamma photons incident at a 45° angle was improved from 2.0 ⋅ 103 to 448 µm FWHM. We conclude that the multi-scale algorithm significantly improves CCD-based gamma cameras as can be applied in future SPECT systems.

Gendered Practices and Landscapes in the Andes: The Shape of Asymmetrical Exchanges

this article describes practices and relations of farming, herding, and cooking that produce and reproduce people and places in culture-specific ways in one region of the central Andes. It also explores how these practices have been changing in relation to regional and global processes surrounding agricultural modernization. the study begins with a look at the degradation of steep slopes and the reduced productivity and social value of women who manage these slopes for small livestock grazing and fuel wood collection. Starting with an ethnographic exploration of local practices and relations of difference, the scope widens to encompass asymmetrical relations of exchange at play in markets, migrations, and development projects, and to consider political decisions and policies that contribute to the uneven terrain on which these exchanges take place. implications for environmental management and conservation include methodological options for approaching environmental problems as integrally social and ecological and for considering these problems in multiscale frames of reference that allow us to examine links among local phenomena and regional or global processes.

Multiscale Dislocation Dynamics Plasticity

A recently developed discrete dislocation dynamics (DD) model for crystalline materials coupled with finite elements (FE) analysis is reviewed. The three-dimensional continuum-based FE formulation for elastoviscoplasticity incorporates the DD simulation, replacing the usual plasticity constitutive relationships, leading to what is called multiscale dislocation dynamics plasticity (MDDP). The coupling involves a nontrivial homogenization to obtain local plastic strains from the contributions of discrete plastic events captured in DD. The superposition principle is used in order to find the effects of the boundaries (free, rigid, or interfaces) on the dislocation movement. The developed computer code can efficiently handle size-dependent small-scale plasticity phenomena and related material instabilities at various length scales ranging from the nano-microscale to the mesoscale. The DD modeling is based on the fundamental physical laws governing dislocation motions and their interactions with various defects, interfaces, and external loadings. The multiscale frame of consideration merges the two scales of nano-microscale, where plasticity is determined, and the continuum scale, where the energy transport is based. In order to illustrate the usefulness of this approach in investigating a wide range of plasticity phenomena, results for a set of case studies are presented. This includes the deformation and dislocation structure during nano-indentation in bcc and fcc single crystals, analyses pertaining to the formation of dislocation boundaries during heavy deformation, dislocations interaction with shock-waves during impact loading conditions, and dislocation–defect interaction.