The wavelet transform is a popular signal processing technique, particularly due to its impressive results in data compression. Its usefulness includes two-dimensional data for use in imageprocessing and three-dimens...
详细信息
ISBN:
(纸本)081942143X
The wavelet transform is a popular signal processing technique, particularly due to its impressive results in data compression. Its usefulness includes two-dimensional data for use in imageprocessing and three-dimensional data for use in video processing. In imageprocessing, the current trends are for image sizes which require a substantial amount of computing power;an application processing a 1024 by 1024 standard quality image requires many millions of processing steps per image frame. When processing sequences of these images for video, the throughput required is considerable in order to attain even low display rates. Three-based architectures have been proposed to provide this throughput rate by processing pixels in a data parallel fashion. Each level of the wavelet transform is processed using an array or a plane of processing elements operating in parallel on shared or distributed data. The largest of these architectures, the plane-based H-tree design, provides a real-time, pipelineable implementation of the 2DWT, but is costly in terms of VLSI area due to its requirement of O(n2) processors for a n by n data-set. In this paper, we look at methods for improving the practicality of these architectures by reducing the required area for a given problem size. This is achieved by adding extra processors at the root of the tree, which allows processing of larger images with an insignificant addition of hardware in exchange for a detrimental effect on the processing speed. We conclude the paper by presenting area/time trade-offs which can be used to evaluate cost/performance specifications.
The problem of mis-registration of 3D datasets is addressed for MR imaging of the breast. Three methods are examined and qualitative evidence of their performance is presented. A method of constrained principal axis d...
详细信息
ISBN:
(纸本)0819420859
The problem of mis-registration of 3D datasets is addressed for MR imaging of the breast. Three methods are examined and qualitative evidence of their performance is presented. A method of constrained principal axis determination is compared with a correlation method, and an orthogonal polynomial approach is extended into three dimensions. The most effective methods still remain relatively time consuming, and for the correlation method, paralleldistributed computing has been applied to reduce the time-to-solution.
The proceedings contain 26 papers. The special focus in this conference is on parallel I/O and parallel Hardware. The topics include: Flexibility and performance of parallel file systems;study of data distribution str...
ISBN:
(纸本)3540616950
The proceedings contain 26 papers. The special focus in this conference is on parallel I/O and parallel Hardware. The topics include: Flexibility and performance of parallel file systems;study of data distribution strategies for parallel I/O management;an experimental evaluation of the parallel I/O systems of the IBM SP and intel paragon using a production application;a scalable interconnection structure for distributed computing;performance aspects of virtual circuit connections over a local ATM network;new results of the salzburg NTN-method for the radon transform;parallel computation of optimal parameters for pseudo random number generation;parallel evaluation of multi-join queries;skew handling in the DBS3 parallel database system;dynamic declustering methods for parallel grid files;encapsulation of intra-operation parallelism in a parallel match operator;synthetic workload generation for parallelprocessing systems;parallel program visualization with MUCH;heuristics to optimize the speed-up of parallel programs;vectorization of visualization algorithms;parallel chaotic iterative algorithms for image reconstruction with limited projection data;the parastation project;a new language for automatic data generation;horizontal fragmentation in distributed object database systems;implementing snapshot protocol for message based communication libraries;a data structure for data dependence analysis in programs with pointers;a partition method for solving block pentadiagonal linear system on intel hypercube iPSC/860 and parallelimageprocessing by using homogeneous computing structures.
The proceedings contain 76 papers. The special focus in this conference is on Applied parallel Computing Industrial Computation and Optimization. The topics include: Performance tuning IBM RS/6000 POWER2 systenls;digi...
ISBN:
(纸本)3540620958
The proceedings contain 76 papers. The special focus in this conference is on Applied parallel Computing Industrial Computation and Optimization. The topics include: Performance tuning IBM RS/6000 POWER2 systenls;digital imageprocessing;use of parallel computations for account of options by a monte carlo method;wavelets and differential equations;from first princlples to industrial applications;parallel implementation of a schwarz domain decomposition algorithm;practical experience in the dangers of heterogeneous computing;coupling the advection and the chemical parts of large air pollution models;advanced optimizations for parallel irregular out-of-core programs;a software architecture for massively parallel input-output;spatial genetic algorithm and its parallel implementation;addressing algebra as a tool for parallel program development;optimizing the NAS BT application for the POWER CHALLENGEarray;providing access to high performance computing technologies;parallel object-oriented library of genetic algorithms;monitoring of performance of PVM applications on virtual network computer;a parallel version of the quasi-minimal residual method based on coupled two-term recurrences;a proposal for parallel sparse BLAS;parallel search-based methods in optimization;an hierarchical approach for performance analysis of scaLAPACK-based routines using the distributed linear algebra machine;scaLAPACK tutorial;bulk synchronous parallelisation of genetic programming;frontal software for the solution of sparse linear equations;parallel heuristics for bandwidth reduction of sparse matrices with IBM SP2 and cray T3D;parallel libraries on distributed memory architectures;parallel least squares estimates of 2-D spect image reconstructions on the SGI power challenge and the prospect for parallel computing in the oil industry.
This paper presents an approach to texture segmentation by thresholding based on compactness measures of fuzzy sets of determine thresholds of an ill-defined image. The extension of fuzziness in the texture feature sp...
详细信息
This paper presents an approach to texture segmentation by thresholding based on compactness measures of fuzzy sets of determine thresholds of an ill-defined image. The extension of fuzziness in the texture feature space provides more meaningful results than by considering fuzziness in gray scale domain. The effectiveness of the algorithm is demonstrated by comparison with other traditional non-fuzzy methods or the controversial fuzzy method in gray scale alone. In addition, the efficiency of our algorithm is further improved by parallel implementation using distributed shared memory workstations.
Many imageprocessing tasks can be computed efficiently in a single program multiple data (SPMD) fashion on massively parallel systems. Although executing SPMD tasks on coarse-grained heterogeneous systems yields a co...
详细信息
Many imageprocessing tasks can be computed efficiently in a single program multiple data (SPMD) fashion on massively parallel systems. Although executing SPMD tasks on coarse-grained heterogeneous systems yields a cost-effective solution, heterogeneity introduces more complexity in data partitioning, mapping, and scheduling problems. In this paper, three image data partitioning schemes for parallelimageprocessing in heterogeneous systems are investigated and implemented using the parallel virtual machine message passing library. The partitioning schemes are based on the system characteristics (processing capability) incorporated within the distributed computing primitives (DCP) environment using SpecInt92 benchmark and our DCP-based benchmark. We compare the results with the baseline (Eq-based) scheme that equally partitions images regardless of processing speed. The results from the experiments show that the DCP-based partitioning scheme outperforms the Eq-based and Spec-based schemes.
This paper presents an approach to texture segmentation by thresholding based on compactness measures of fuzzy sets to determine thresholds of an ill-defined image. The extension of fuzziness in the texture feature sp...
详细信息
This paper presents an approach to texture segmentation by thresholding based on compactness measures of fuzzy sets to determine thresholds of an ill-defined image. The extension of fuzziness in the texture feature space provides more meaningful results than by considering fuzziness in gray scale *** effectiveness of the algorithm is demonstrated by comparison with other traditional nonfuzzy methods or the controversial fuzzy method in gray scale *** addition,the efficiency of our algorithm is further improved by parallel implementation using distributed shared memory workstations.
Hypermedia systems are very important as platforms in the future multimedia information network services. In order to provide more flexible and sophisticated multimedia presentation, it is desirable to simply and effi...
详细信息
Hypermedia systems are very important as platforms in the future multimedia information network services. In order to provide more flexible and sophisticated multimedia presentation, it is desirable to simply and efficiently transmit various media data from the databases distributed over network and present them on the user station depending on scenario so as to semantically integrate into multimedia information objects. In this paper, a new hypermedia system which is based on object-oriented structure to perform two different temporal synchronization methods, "lip synchronization" and "scene synchronization" is introduced. Lip synchronization method performs, fine-grained temporal synchronization between continuous media, such as audio and video. On the other hand in scene synchronization, coarse-grained temporal relation between continuous media and discrete media, such as image, graphics or text is maintained depending on the presentation scenario. Furthermore, several media which are semantically related to each other are unified into a Media Group (MG) to simplify its media treatment. The MGs and their involved processing functions are encapsulated into a Media Object (MO) and easily controlled by message passing procedure. In this paper, implementation of the hypermedia model which performs scene synchronization function between discrete and continuous media in the MOs depending on the presentation scenario, and context switching function between each media presentation is precisely described. Finally "Electric Museum" as a prototype application of hypermedia systems is introduced to evaluate the suggested synchronization mechanism.
This paper aims to develop effective parallelprocessing techniques for 3-dimensional (3D) multi-level median filtering (3DMMF) with motion compensation on video sequences. Due to the nature of unbalanced load of the ...
详细信息
ISBN:
(纸本)0780331222
This paper aims to develop effective parallelprocessing techniques for 3-dimensional (3D) multi-level median filtering (3DMMF) with motion compensation on video sequences. Due to the nature of unbalanced load of the filtering algorithm and the objective for the parallel approach, two methods of dynamic load balancing have been proposed and compared. They are sender-initiated-load-balancing (SILB) and receiver-initiated-load-balancing (RILB) algorithms. We propose a SILB algorithm which utilises the spatial-temporal characteristics of the processed sequences for load prediction to achieve dynamic load balancing. Both theoretical analysis and experimental results on the IBM SP2 computing surface have been presented in this paper.
This paper presents a parallel approach to a hierarchical image matching scheme using the Hausdorff distance for object recognition and localization in aerial images. Unlike the conventional matching methods in which ...
详细信息
This paper presents a parallel approach to a hierarchical image matching scheme using the Hausdorff distance for object recognition and localization in aerial images. Unlike the conventional matching methods in which edge pixels are considered as image feature pixels in distance transform and the blind pointwise comparison procedure in terms of the Hausdorff distance is applied, a guided image matching system is developed by the hierarchical detection of interesting points via a dynamic thresholding scheme for the search of the best matching between two image sets. Furthermore, the concept of remote procedure call (RPC) in distributed systems is introduced for the parallel implementation to achieve the speedup without specific software and hardware requirement.
暂无评论