Hound:a parallel image distribution system for cluster based on Docker

Current applications,consisting of multiple replicas,are packaged into lightweight containers with their execution dependencies.Considering the dominant impact of distribution efficiency of gigantic images on container startup(e.g.,distributed deep learning application),the image“warm-up”technique which prefetches images of these replicas to destination nodes in the cluster is proposed.However,the current image“warm-up”technique solely focuses on identical image distribution,which fails to take effect when distributing different images to destination nodes.To address this problem,this paper proposes Hound,a simple but efficient cluster image distribution system based on Docker.To support diverse image distribution requests of cluster nodes,Hound additionally adopts node-level parallelism(i.e.,downloading images to destination nodes in parallel)to further improve the efficiency of image distribution.The experimental results demonstrate Hound outperforms Docker,kubernetes container runtime interface(CRI-O),and Docker-compose in terms of image distribution performance when cluster nodes request different images.Moreover,the high scalability of Hound is evaluated in the scenario of ten nodes.

Development of a mass model in estimating weight-wise particle size distribution using digital image processing

Particle size distribution of coarse aggregates through mechanical sieving gives results in terms of cumu- lative mass percent. But digital image processing generated size distribution of particles, while being fast and accurate, is often expressed in terms of area function or number of particles. In this paper, a mass model is developed which converts the image obtained size distribution to mass-wise distribution, mak- ing it readily comparable to mechanical sieving data. The concept of weight/particle ratio is introduced for mass reconstruction from 2D images of particle aggregates. Using this mass model, the effects of several particle shape parameters （such as major axis, minor axis, and equivalent diameter） on sieve-size of the particles is studied. It is shown that the sieve-size of a particle strongly depend upon the shape param- eters, 91% of its variation being explained by major axis, minor axis, bounding box length and equivalent diameter. Furthermore, minor axis gives an overall accurate estimate of particle sieve-size, error in mean size （D-50） being just 0.4%. However, sieve-size of smaller particles （〈20 ram） strongly depends upon the length of the smaller arm of the bounding box enclosing them and sieve-sizes of larger particles （〉20 mm） are highly correlated to their equivalent diameters. Multiple linear regression analysis has been used to generate overall mass-wise particle size distribution, considering the influences of all these shape parameters on particle sieve-size. Multiple linear regression generated overall mass-wise particle size distribution shows a strong correlation with sieve generated data. The adjusted R-square value of the regression analysis is found to be 99 percent （w.r,t cumulative frequency）. The method proposed in this paper provides a time-efficient way of producing accurate （up to 99%） mass-wise PSD using digital image processing and it can be used effectively to renlace the mechanical sieving.

Modeling random telegraph signal noise in CMOS image sensor under low light based on binomial distribution

The random telegraph signal noise in the pixel source follower MOSFET is the principle component of the noise in the CMOS image sensor under low light. In this paper, the physical and statistical model of the random telegraph signal noise in the pixel source follower based on the binomial distribution is set up. The number of electrons captured or released by the oxide traps in the unit time is described as the random variables which obey the binomial distribution. As a result,the output states and the corresponding probabilities of the first and the second samples of the correlated double sampling circuit are acquired. The standard deviation of the output states after the correlated double sampling circuit can be obtained accordingly. In the simulation section, one hundred thousand samples of the source follower MOSFET have been simulated,and the simulation results show that the proposed model has the similar statistical characteristics with the existing models under the effect of the channel length and the density of the oxide trap. Moreover, the noise histogram of the proposed model has been evaluated at different environmental temperatures.

Fault Diagnosis for IC Engines Using Wavelet Packet and Image Processing

There are few applications of image processing technology for diagnosing andstate monitoring for internal combustion (IC) engines, which is discussed in detail in this paper.The time-frequency distribution images of cylinder head vibration signals are obtained bydecomposing them with a wavelet packet algorithm. It is the first time that we look attime-frequency distribution images from the point of images. Based on this, a new method forapplying image processing technology for diagnosing and state monitoring for internal combustionengines is presented in this paper. A valve fault diagnosis model is set up by image matching, whichis realized on a four-stroke, six-cylinder diesel engine. At the same time, some notes arepresented in this paper. It has been proved that it is of no good effect to diagnose with histogramsof time-frequency images generated by cylinder head vibration signals that have been processed witha wavelet packet algorithm. The reason is given in this paper. Comparisons of diagnosing effect arecarried out between noise-added signals and original signals. It has little effect on diagnosingresults after signals have been added with noise. The results show that this method has a clearphysical meaning and is of good engineering practicability, feasibility, good precision and highspeed.

An improved particle tracking velocimetry(PTV) technique to evaluate the velocity field of saltating particles

Velocity is a key parameter characterizing the movement of saltating particles. High-speed photography is an efficient method to record the velocity. But, manually determining the relevant information from these photographs is quite laborious. However, particle tracking velocimetry（PTV） can be used to measure the instantaneous velocity in fluids using tracer particles. The tracer particles have three basic features in fluids： similar movement patterns within a small region, a uniform particle distribution, and high particle density. Unfortunately, the saltation of sand particles in air is a stochastic process, and PTV has not yet been able to accurately determine the velocity field in a cloud of blowing sand. The aim of the present study was to develop an improved PTV technique to measure the downwind（horizontal） and vertical velocities of saltating sand. To demonstrate the feasibility of this new technique, we used it to investigate two-dimensional saltation of particles above a loose sand surface in a wind tunnel. We analyzed the properties of the saltating particles, including the probability distribution of particle velocity, variations in the mean velocity as a function of height, and particle turbulence. By automating much of the analysis, the improved PTV method can satisfy the requirement for a large sample size and can measure the velocity field of blowing sand more accurately than previously-used techniques. The results shed new light on the complicated mechanisms involved in sand saltation.

Particle image velocimetry measurement of velocity distribution at inlet duct of waterjet self-propelled ship model

A vehicle-mounted three-dimensional underwater particle image velocimetry（PIV） device is used in a towing tank to measure the velocity distribution of the inlet duct of a waterjet ship model in a self-propulsion test. The following points are shown through a comparison of the influences of the stationary and free states of the ship model on the measured results：（1） during the test, the ship attitude will change, specifically, the ship model will heave and trim,（2） the degree of freedom disturbs the processing of the pixel images enough to distort the subsequent image processing,（3） the stationary state of the ship model is the optimal mode for measuring the velocity distribution using the PIV device, and（4） if the changes must be considered, the man-made heaving and trimming may be pre-applied, and be made a corrected stationary mode. In addition, the momentum effect coefficient and the energy effect coefficient are calculated in a non-uniform inflowing state, and the related factors affecting the two coefficients are analyzed. The test results show that the pumping action of the waterjet creates a transverse vector in the cross-sectional speed, which increases the non-uniformity of the inflow. These results could help to establish the design requirements for a waterjet-propelled ship type.

Three-dimensional free view reconstruction in axially distributed image sensing

Three-dimensional （3D） acquisition has been the most important technique for augmented reality or virtual reality industries. Stereo vision can obtain 3D infor- mation of objects using disparity between two images captured by a stereo camera. However, due to lack of per- spectives for 3D objects in this technique, accurate 3D information may not be generated and the number of viewing points is limited. By contrast, multi-view or super multi-view imaging techniques can obtain more accurate 3D information because they use multiple cameras for pickup of 3D objects so that more perspec- tives and viewing points of 3D objects can be recorded.

On-line full scan inspection of particle size and shape using digital image processing

An on-line full scan inspection system is developed for particle size analysis. A particle image is first obtained through optical line scan technology and is then analyzed using digital image processing. The system is composed of a particle separation module, an image acquisition module, an image processing module, and an electric control module. Experiments are carried out using non-uniform 0.1 mm particles. The main advantage of this system consists of a full analysis of particles without any overlap or miss, thus improving the Area Scan Charge Coupled Device （CCD） acquisition problems. Particle size distribution, roundness, and sphericity can be obtained using the system with a deviation of repeated precision of around ±1%. The developed system is shown to be also convenient and versatile for any particle size and shape for academic and industrial users.

Measurement, modelling, and closed-loop control of crystal shape distribution： Literature review and future perspectives

Crystal morphology is known to be of great importance to the end-use properties of crystal products, and to affect down-stream processing such as filtration and drying. However, it has been previously regarded as too challenging to achieve automatic closed-loop control. Previous work has focused on controlling the crystal size distribution, where the size of a crystal is often defined as the diameter of a sphere that has the same volume as the crystal. This paper reviews the new advances in morphological population balance models for modelling and simulating the crystal shape distribution （CShD）, measuring and estimating crystal facet growth kinetics, and two- and three-dimensional imaging for on-line characterisation of the crystal morphology and CShD. A framework is presented that integrates the various components to achieve the ultimate objective of model-based closed-loop control of the CShD. The knowledge gaps and challenges that require further research are also identified.

Development of 3-actoxymethoxycarbonyl-2,2,5,5-tetramethyl-1-pyrrolidinyloxyl as an electron paramagnetic resonance imaging reagent for in vivo mapping brain oxygen distribution and infarction in ischemic brain

Measurement of oxygen concentration and distribution in brain is essential to understanding the pathophysiology of stroke. Although brain oxygen level is critical for brain tissue survival,

NanoSIMS analytical technique and its applications in earth sciences

Despite the significant improvement on spatial resolution, NanoSIMS still preserves relatively high mass resolution, sensitivity, and analytical precision. It has become an important analytical platform to determine chemical compositions of solid materials, and has been widely used in space, earth, life, and materials sciences, etc. By using a Cs＋ ion beam with a size as small as 50 nm scanning over sample surfaces, we are able to obtain high spatial resolution images of up to 7 species simultaneously. When utilizing Faraday cup, high analytical precision of 0.3‰-0.5‰ （1SD） for C, O and S isotopic analysis can be achieved. Although this precision level is still lower than that of conventional SIMS, it already meets the major requirements of Earth Sciences. In 2011, the first NanoSIMS of China （Cameca NanoSIMS 50L） was installed at Institute of Geology and Geophysics, Chinese Academy of Sciences. Based on the working mechanism and analytical modes of the instrument, this paper will systematically introduce the analytical methods established with the NanoSIMS and their potential applications in earth sciences. These methods include trace element distribution images in mineral zoning, high spatial resolution （2-5/am） Pb-Pb and U-Pb dating, water content and H isotopic analysis for silicate glass and apatite, C isotopic analysis for diamond and graphite, O isotopic analysis for carbonate, S isotopic analysis for sulfides. In addition, the specific requirements for sample preparation will also be introduced in order to facilitate domestic earth scientists＇ use.

Real-time Follow-up Head Tracking in Dynamic Complex Environments

In the modes of both object motion and camera motion,an enhanced Camshift algorithm,which is based on suppressing similar color features of background and on joint color probability density distribution image,is proposed to real-time track head in dynamic complex environment.The system consists of face detection module,head tracking module and camera control module.When tracking fails,a self-recovery mechanism is introduced.At first the Adaboost face detector based on Haar-like features is implemented to find frontal faces,the false positive is filtered according to the skin color criterion,and the true face is used to initialize the tracking module.In hue saturation value(HSV) colorspace,the hue-saturation(H-S) histogram of face skin and the saturation-value(S-V) histogram of hair are built to produce the joint color probability density distribution image,and this is intended to realize the head tracking with arbitrary pose.During tracking,region of interest(ROI) is introduced,and the color probability density distribution of a specified background area outside the ROI is learned,similar color features in the head are suppressed according to the learning result.The background suppression step is intended to resolve the problem that the tracker maybe fails when the head is distracted by backgrounds having similar colors with the head.A closed loop control model based on speed regulation is applied to drive an active camera to center the head.Once tracking drift or failure is detected,the system stops tracking and returns to the face detection module.Our experimental results show that the presented system is well suitable for tracking head with arbitrary pose in dynamic complex environments,also the active camera can track moving head smoothly and stably.The system is computationally efficient and can run in real-time completely.