检索结果-内蒙古大学图书馆

Performance analysis of parallel high-resolution image restoration algorithms on Intel supercomputer

CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE 2021年第4期33卷 e5996-e5996页

作者： Lirkov, Ivan Harizanov, Stanislav Paprzycki, Marcin Ganzha, Maria Bulgarian Acad Sci Inst Informat & Commun Technol Sofia 1113 Bulgaria Bulgarian Acad Sci Inst Math & Informat Sofia Bulgaria Polish Acad Sci Syst Res Inst Dept Intelligent Syst Warsaw Poland Warsaw Univ Technol Dept Math & Informat Sci Warsaw Poland

In this article, we present an experimental performance study of a parallel implementation of two Poissonian image restoration algorithms. Hybrid parallelization, based on MPI and OpenMP standards, is investigated. The implementation is tested for high-resolution radiographic images, on a supercomputer based on Intel Xeon processors, combined with Intel Xeon Phi coprocessors. The experimental results show an essential improvement in the execution times, when running experiments for a variety of problem sizes, and number of threads.

关键词： Anscombe transform image restoration parallel algorithm Intel Xeon Phi coprocessor

来源：评论

学校读者我要写书评

暂无评论

A PANALLEL ALCONRRHM RPNRCE ASIAN OPTIONS WITH MULTI-DTMENSTONAL ASSETS

A PANALLEL ALCONRRHM RPNRCE ASIAN OPTIONS WITH MULTI-DTMENST...

引用

作者： Kai Huang University of Manitoba

学位级别：硕士

Options are no\M traded actively on many exchanges across the world. Especially, some complex options such as multi-dimensional Asian options are much more widely used in the financial market. The price of an option need to be computed to help investors in deciding if it is worthwhile entering a particular contract. Pricing multldimensional American-style Asian options is difficult because it involves early exercise possibilities, multi-dimensional asset management, as well as the path-dependency issue. Meanwhile, study of pricing multi-dimensional American-style Asian options is critical since it is widely traded in the market place. We address this problem in the current study. The binomial tree method is one of the traditional numerical techniques that has been used for pricing options in the financial community. Due to the size and complexity of the options traded these days, this method is slow. However, this method is amenable to parallel implementation to make it faster. In this thesis, we have developed a parallel algorithm for pricing the American-style Asian options and improved this algorithm by introducing a mathematical transformation to handle the interrelation of multi-dimensionality of the problem as well as reduce the computational cost to deal with the multi-dimensionality and hence can lead to quicker and accurate solutions. We have tested the new algorithm on a 16 nodes distributed computing system and compared the new algorithm with the parallel algorithm for American-style plain op- tions. We have also provided analytical results for the computation and communication complexities of the parallel algorithm. Option values are computed and the experimental results show that the algorithm converges. The mathematical transformation is proved to be essential through our experimental results. Compared with the parallel algorithm for pricing American-style plain options, the algorithm for pricing American-style Asian options offers a better problem for

关键词： Financial Derivatives Option Pricing Asian Options parallel algorithm

来源：评论

学校读者我要写书评

暂无评论

Three-dimensional acoustic parabolic equation model based on GPU processing,2 23

Three-dimensional acoustic parabolic equation model based on...

引用

23rd International Congress on Acoustics: Integrating 4th EAA Euroregio, ICA 2019

作者： Lee, Keunhwa Seong, Woojae Na, Youngnam Sejong University Korea Republic of Seoul National University Korea Republic of Agency for the Defense Development Korea Republic of

ISBN: (纸本)9783939296157

We introduce a three-dimensional acoustic parabolic equation (PE) model based on GPU processing. This model is developed by splitting the full exponential operator in the three-dimensional split-step marching solution in the Cartesian coordinate. Each exponential operator is approximated by the Taylor series and the sum representation of Padé approximation, which gives an opportunity for GPU parallel processing of the 3D PE solver. An advantage of this model is to implement higher-order cross-terms appearing to the operator approximation efficiently. In the talk, we will show the computational performance of GPU based PE code with the comparison of a traditional code for several ocean environments. © 2019 Proceedings of the International Congress on Acoustics. All rights reserved.

关键词： GPU parallel algorithm Three-dimensional parabolic acoustic equation (3DPE)

来源：评论

学校读者我要写书评

暂无评论

Data assimilation for large-scale computational models 54

Data assimilation for large-scale computational models

引用

54th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference

作者： Khalil, Mohammad Subber, Waad Sarkar, Abhijit Department of Civil and Environmental Engineering Mackenzie Building Carleton University Colonel By Drive Ottawa ON K1S 5B6 Canada

来源：评论

学校读者我要写书评

暂无评论

The ultimate bead counter: efficient massively parallel simulation of ballistic deposition 95

The ultimate bead counter: efficient massively parallel simu...

引用

Proceedings of the ninth workshop on parallel and distributed simulation

作者： Boris D. Lubachevsky Vladimir Privman Subhas C. Roy AT&T Bell Laboratories 600 Mountain Avenue Murray Hill NJ Department of Physics Clarkson Univercity Potsdam NY Department of Computer Science College of William and Mary Williamsburg VA

ISBN: (纸本)9780818671203

We simulate ballistic particle deposition wherein a large number of spherical particles are “dropped” vertically over a planar horizontal surface. Upon first contact (with the surface or with a previously deposited particle) each particle stops. This model helps material scientists to study the adsorption and sediment formation [1]. The model is sequential, with particles deposited one by one. We have found an equivalent formulation using a continuous time random process and we simulate the latter in parallel using a method similar to the one previously employed for simulating Ising spins [2]. We augment the parallel algorithm for simulating Ising spins with several techniques aimed at the increase of efficiency of producing the particle configuration and statistics collection. Some of these techniques are similar to [3], [4], [5]. We implement the resulting algorithm on a 16K PE MasPar MP-1 and a 4K PE MasPar MP-2. The parallel code runs on MasPar computers two orders of magnitude faster than an optimized sequential code runs on a fast workstation.

关键词： massively parallel simulation 16K PE MasPar MP-1 parallel algorithm Ising model digital simulation atom-surface impact parallel algorithms statistics collection physics computing particle configuration equivalent formulation ultimate bead counter material scientists Ising spins simulation 4K PE MasPar MP-2 continuous time random process ballistic deposition sediment formation adsorption sputter deposition

来源：评论

学校读者我要写书评

暂无评论

Clutter cancellation in passive radar from the perspective of maximum likelihood

Clutter cancellation in passive radar from the perspective o...

引用

2021 CIE International Conference on Radar, Radar 2021

作者： Lyu, Xiaoyong Xu, Jin Wang, Jun School of Information Engineering Zhengzhou University Zhengzhou China The 38th Research Institute of China Electronics Technology Group Corporation Hefei China Xidian University National Laboratory of Radar Signal Processing Xi'an China

ISBN: (纸本)9781665498142

Clutter cancellation in passive radar is gaining increasing interest. The extensive cancellation algorithm (ECA) has received a range of applications in passive radar clutter cancellation. However, the ECA suffers from great computational complexity. Several variants of ECA have been proposed to increase the computational efficiency. Among those methods, a parallel implementation algorithm obtains an interesting conclusion that, the cancellation weight vector of ECA can be calculated as the weighted mean of several weight vectors, each of which is computed using only a small segment of the reference and surveillance signals. The algorithm can get very similar cancellation weight vector compared with the ECA, but increases the computational efficiency. However, the derivation of this algorithm in the original paper is very complicated and not straightforward to be understood and the obtaining of the final conclusion is not explicit. In this paper, we rederive this algorithm from the maximum likelihood (ML) perspective. We can get the same algorithm, but use a very simple and straightforward derivation. Furthermore, we show explicitly how the finial conclusion is derived, by making some assumptions on the statistical characteristics of the illuminating signal, which is not discussed in the original paper. Finally, we analyze through simulations how the assumptions made on the illuminating signal affect the estimation of the cancellation weight vector. © 2021 IEEE.

关键词： clutter cancellation maximum likelihood parallel algorithm passive radar

来源：评论

学校读者我要写书评

暂无评论

The Non-eye Typhoon Center Locating with parallel Computing

The Non-eye Typhoon Center Locating with Parallel Computing

引用

Proceedings of The Fourth International Conference on parallel and Distribyted Computing,Applications and Technologies(第四届并行与分布式计算应用与技术国际会议)

作者： Juntao Xue Zhengguang Liu Bing Wu Huanzhu Liu School of Engineering Electrical and Automation Tianjin University 300072 China National Meteorological Center Beijing 100081 China

ISBN: (纸本)0780378407

A new mathematical morphology-based algorithm is proposed to achieve automatic center location of non-eye typhoon. The center of a non-eye typhoon is near the geometric center of the cloud system and has higher temperature. For each infrared satellite cloud image, the locating procedures are as follows: a) noises filtering, b) main cloud systems segmenting, c) center locating and d) multispectral image verification. The algorithms are based on operations of mathematical morphology, and designed in IBM SP2 massively parallel computer. The experiment results show that the algorithm locates the centers of most non-eye typhoons successfully and achieves faster, more precise and non-human interactive non-eye typhoon center positioning.

关键词： Satellite cloud image Non-eye typhoon locating parallel algorithm Mathematical morphology

来源：评论

学校读者我要写书评

暂无评论

Efficient Viewshed Computation algorithms on GPUs and CPUs

Efficient Viewshed Computation Algorithms on GPUs and CPUs

引用

作者： Qarah, Faisal F. University of South Florida

学位级别：Ph.D.

Nowadays with the advance in managing and collecting large data, GIS is one of the applications that suffer from lack of efficient data management methods. GIS data often come in form of maps with different types of data such as temperature, topology, and population. This dissertation focuses on exact-viewsheds computation for large terrains, and due to the poor performance of current exact-viewshed algorithms that may need several hours to process a midsize map, we found the need for new algorithms that are capable of efficiently computing viewshed for large size maps. This work presents a highly-efficient exact-viewshed computation algorithm based on the radial-sweep algorithm, implemented and optimized for GPUs. The first version of our GPU algorithm shows significant improvement in performance over the sequential CPU-based algorithm, providing at least an order of magnitude speedup. We further improve the algorithms by tackling two challenges in parallelizing the radial-sweep algorithm: (1) dividing the elevation-grid into small-size sectors accurately between threads and (2) consumption of excessive amount of memory. Both problems have been resolved in the newly proposed algorithms. We were able to parallelize the algorithm by building a histogram for cells’ events based on their azimuth-angle in the grid. The histogram we help each thread to know the number of events it has to process which forms a small size sector. Then, we applied prefix-sum operation on the histogram so each thread will know where to start processing the events. The second problem which is the excessive memory consumption has been resolved by dividing the grid into sectors and processing them individually to obtain a smaller intermediate result instead of processing the whole grid that would yield a larger intermediate result. Also, to search for cell IDs in each sector, the grid is divided into scan-regions where the overhead of the scan operation is minimal. The new GPU-based algorithm i

关键词： Computing visibility GIS GPU parallel algorithm Spatial data analysis

来源：评论

学校读者我要写书评

暂无评论

Retina identification based on moment invariant 122

Retina identification based on moment invariant

引用

2015 122nd ASEE Annual Conference and Exposition

作者： Kavianpour, Alireza Shoari, Simin Kavianpour, Behdad DeVry University PomonaCA United States University of California Gavin Herbert Eye Institute Irvine United States

This paper demonstrates the importance of topics in physics and mathematics courses such as matrix, eigenvectors, centroid, and moment of inertia in the pattern recognition applications. Teaching advance topics in physics and mathematics is not an easy task. Students always ask this question: What is the use? The best practice for teaching these topics is to combine them with a real industry application. For example, computer hard disk reading is based on the concept of derivatives and the change in the magnetic fields. To demonstrate this type of teaching, the use of three dimensional moments in the pattern recognition is explained. The three-dimensional moments may be used to detect different patterns in a digitally represented image. The combination of several moments for an object has an invariant property. The mathematical foundation of an invariant feature is related to the theory that images taken from different angles from the same object have the same set of moment invariants. Moments contain information of an image which can be used in calculating the location and orientation of an object. An algorithm for recognition of an individual identity based on a digitally represented image of the scanned retina is presented. The technology is based upon the fact that no two retinal patterns are alike. In this paper, the nine parameters of an ellipsoidal shape fitted into a retinal image such as coordinates of the center of the ellipsoid, the length of major, minor, intermediate axes, and the direction of three axes will be calculated. For each individual, these parameters are unique. © American Society for Engineering Education, 2015.

关键词： Ellipsoid Image processing parallel algorithm Pattern recognition Retina Three-dimensional moments

来源：评论

学校读者我要写书评

暂无评论

Rank order filtering on bit-serial mesh-connected computers

Rank order filtering on bit-serial mesh-connected computers

引用

Conference on Statistical and Stochastic Methods for Image Processing

作者： Shi, HC Li, HZ UNIV MISSOURI DEPT COMP SCI & COMP ENGNCOLUMBIAMO 65211

ISBN: (纸本)0819422118

Rank order filters have a wide variety of applications in image processing as an effective tool for removing noise in images without severely distorting abrupt changes in the images. The k-th rank filter with an m X m window sets each pixel the k-th smallest value of the m² pixels in its m X m neighborhood. The median filter is the most commonly used special case of rank order filters. Rank order filtering requires intensive computation. In this paper, we consider implementation of rank order filters on bit-serial mesh-connected computers such as the Lockheed Martin CISP computer. We design rank filtering algorithms using the threshold decomposition and radix splitting techniques, and present some experimental results of the implementation of those algorithms on the CISP computer.

关键词： rank order filter threshold decomposition radix splitting parallel algorithm mesh-connected computer

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：