Key technique research on desktop virtualization in cloud environment

Desktop Virtualization is an essential problem in cloud manufacturing to make full use of shared resources so as to meet the practical application requirements.In this paper,an efficient virtualization solution for desktop transferring is presented for sharing resources.Our strategy is not limited to transfer desktops between virtual machines(VMs),transmission between virtual and physical machines is also achieved.A three-layer systemic structure based on web environment is proposed in the study.We employ the application program library Libvirt and kernel-based virtual machine(KVM)technique to manage VMs with dynamic customization scheme,which gains better virtualization performance.Moreover,we present two different schemes for desktop graphics transferring.One is based on the transport protocol virtual network computing(VNC),and the other adopts the simple protocol for independent computing environment(SPICE),which both achieve more satisfactory results.The validity of our scheme is demonstrated on some experimental results and systemic qualitative evaluation is also provided.

Reconstruction of fluid surface using physical property

A novel technique for reconstructing fluid surface from video is introduced.Both fluid motion vectors as well as Lattice Boltzmann Method(LBM)are employed in the study.Region-based correlation method is used to initialize motion vectors field,after clustering fluid motion vector results can be obtained.Then the height geometry information of fluid surface can be calculated from fluid motion vector further.At last,the distribution of fluid particle is interpolated and the height field can be further refined.Reconstruction results are demonstrated with several challenge videos.The experimental results show that the method is convenient and efficient.The calculation results can reflect the characteristics of the fluid movement,and it is a valid method for reconstructing fluid surface.

REAL TIME RECONSTRUCTION OF FLUID IN VIDEO

This paper puts forward a new method of realtime reconstruction of fluid in natural scene.It takes the measure of combination of image analysis and LBM(Lattice Boltz-mann Methods).First,employs LK(Lucas–Kanade)method to calculate the dense optical flow,and then takes LBM to obtain the joint force of central particles for the initial result.After backfilling the velocity vectors field,it adopts the K-means cluster to obtain several classes,in each class,it takes advantage of the Rayleigh distribution to fit the height field of fluid.Finally,the reconstruction result of fluid is obtained.In addition,it demonstrates the results of the height field of fluid in the experiment.Fur-ther experiments shows that it is a valid method of fluid reconstruction with real time and can be used in the study of natural landscape fluid with efficiency and feasibility.

Physically-based enhancement in shallow water resimulation

To resimulate a customized fluid derived product by analyzing an existing fluid is significant and difficult.This paper proposes a driven model recovery method,which is challenging in fluid resimulation customization.First,fluid physical properties are calculated under the constraints of appearance and dynamic behavior of the example water.Second,a hybrid particle lattice Boltzmann method for shallow water(LBMSW)is recovered from the dynamic geometry on fluid surface.As it is found that the resimulation details fade gradually with LBMSW auto-advection,a physically-based enhancement scheme is presented.A nonlinear algorithm is introduced to stretch the faded density to retain resimulation details.Experiments show that the proposed approach can obtain more realistic resimulation products in several challenging scenarios.