A software architecture for user transparent parallel image processing

Seinstra, F.J.; Koelma, D.C.; Geusebroek, Jan‐Mark

doi:10.1016/s0167-8191(02)00103-5

Cited by 54 publications

(61 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As more MPI processes are added to the configuration the total execution time reduces linearly. These results are entirely in line with earlier speedup characteristics reported in [12] for much larger cluster systems. We would like to stress that these speedup results are obtained without requiring any parallelization effort from the application programmer.…”

Section: Discussionsupporting

confidence: 92%

See 1 more Smart Citation

Towards User Transparent Parallel Multimedia Computing on GPU-Clusters

Werkhoven

Maassen

Seinstra

2011

Computer Architecture

View full text Add to dashboard Cite

Abstract. The research area of Multimedia Content Analysis (MMCA) considers all aspects of the automated extraction of knowledge from multimedia archives and data streams. To satisfy the increasing computational demands of MMCA problems, the use of High Performance Computing (HPC) techniques is essential. As most MMCA researchers are not HPC experts, there is an urgent need for 'familiar' programming models and tools that are both easy to use and efficient. Today, several user transparent library-based parallelization tools exist that aim to satisfy both these requirements. In general, such tools focus on data parallel execution on traditional compute clusters. As of yet, none of these tools also incorporate the use of many-core processors (e.g. GPUs), however. While traditional clusters are now being transformed into GPU-clusters, programming complexity vastly increases -and the need for easy and efficient programming models is as urgent as ever. This paper presents our first steps in the direction of obtaining a user transparent programming model for data parallel and hierarchical multimedia computing on GPU-clusters. The model is obtained by extending an existing user transparent parallel programming system (applicable to traditional compute clusters) with a set of CUDA compute kernels. We show our model to be capable of obtaining orders-of-magnitude speed improvements, without requiring any additional effort from the application programmer.

show abstract

Section: Discussionsupporting

confidence: 92%

“…The example is selected because results for data parallel execution with Parallel-Horus are well available, and presented in [12].…”

Section: A Line Detection Applicationmentioning

confidence: 99%

Towards User Transparent Parallel Multimedia Computing on GPU-Clusters

Werkhoven

Maassen

Seinstra

2011

Computer Architecture

View full text Add to dashboard Cite

show abstract

“…For CPU-only execution, the execution time reduces linearly; using 8 CPUs gives a speedup of 7.9. These results are entirely in line with earlier speedup characteristics reported in [44] for much larger cluster systems.…”

Section: Experiments 2: User Transparent Mmca On Gpu-clusterssupporting

confidence: 92%

“…To effectively exploit the available processing power, a thorough understanding of the complexity of such systems is essential. As a consequence, the number of scientists capable of using such systems effectively (if at all) is relatively low [44].…”

Section: Introductionmentioning

confidence: 99%

Jungle Computing: Distributed Supercomputing Beyond Clusters, Grids, and Clouds

Seinstra

Maassen

Nieuwpoort

et al. 2011

Computer Communications and Networks

View full text Add to dashboard Cite

In recent years, the application of high-performance and distributed computing in scientific practice has become increasingly wide-spread. Among the most widely available platforms to scientists are clusters, grids, and cloud systems. Such infrastructures currently are undergoing revolutionary change due to the integration of many-core technologies, providing orders-of-magnitude speed improvements for selected compute kernels. With high-performance and distributed computing systems thus becoming more heterogeneous and hierarchical, programming complexity is vastly increased. Further complexities arise because urgent desire for scalability, and issues including data distribution, software heterogeneity, and ad-hoc hardware availability, commonly force scientists into simultaneous use of multiple platforms (e.g. clusters, grids, and clouds used concurrently). A true computing jungle. In this chapter we explore the possibilities of enabling efficient and transparent use of Jungle Computing Systems in every-day scientific practice. To this end, we discuss the fundamental methodologies required for defining programming models that are tailored to the specific needs of scientific researchers. Importantly, we claim that many of these fundamental methodologies already exist today, as integrated in our Ibis high-performance distributed programming system. We also make a case for the urgent need for easy and efficient Jungle Computing in scientific practice, by exploring a set of state-of-the-art application domains. For one of these domains we present results obtained with Ibis on a real-world Jungle Computing System. The chapter concludes by exploring fundamental research questions to be investigated in the years to come.

show abstract

“…A lot of research was done using various strategies to parallelize the sequential algorithms (Stevenson, et al, 2000) (Jones, et al, 2003) (Akoguz, et al, 2013) (Bhojne, et al, 2013) (Plaza & Chang, 2007). Those various implementations require image processing domain knowledge from scientists and computer knowledge from computer engineers, which creates a gap between image processing and high performance (Seinstra, et al, 2002). The researchers (Seinstra, et al, 2002) initialized a software architecture with an extensive library of data parallel low level image operations (assuming the computing environment to be homogeneous distributed multi-computers).…”

Section: Parallel Computingmentioning

confidence: 99%

High Performance Computing for DSM Extraction From Zy-3 Tri-Stereo Imagery

Huang

Pan

et al. 2016

ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:ZY-3 has been acquiring high quality imagery since its launch in 2012 and its tri-stereo (three-view or three-line-array) imagery has become one of the top choices for extracting DSM (Digital Surface Model) products in China over the past few years. The ZY-3 tristereo sensors offer users the ability to capture imagery over large regions including an entire territory of a country, such as China, resulting in a large volume of ZY-3 tri-stereo scenes which require timely (e.g., near real time) processing, something that is not currently possible using traditional photogrammetry workstations. This paper presents a high performance computing solution which can efficiently and automatically extract DSM products from ZY-3 tri-stereo imagery. The high performance computing solution leverages certain parallel computing technologies to accelerate computation within an individual scene and then deploys a distributed computing technology to increase the overall data throughput in a robust and efficient manner. By taking advantage of the inherent efficiencies within the high performance computing environment, the DSM extraction process can exploit all combinations offered from a set of tri-stereo images (forward-backword, forward-nadir and backword-nadir). The DSM results merged from all of the potential combinations can minimize blunders (e.g., incorrect matches) and also offer the ability to remove potential occlusions which may exist in a single stereo pair, resulting in improved accuracy and quality versus those that are not merged. Accelerated performance is inherent within each of the individual steps of the DSM extraction workflow, including the collection of ground control points and tie points, image bundle adjustment, the creation of epipolar images, and computing elevations. Preliminary experiments over a large area in China have proven that the high performance computing system can generate high quality and accurate DSM products in a rapid manner.

show abstract

A software architecture for user transparent parallel image processing

Cited by 54 publications

References 16 publications

Towards User Transparent Parallel Multimedia Computing on GPU-Clusters

Towards User Transparent Parallel Multimedia Computing on GPU-Clusters

Jungle Computing: Distributed Supercomputing Beyond Clusters, Grids, and Clouds

High Performance Computing for DSM Extraction From Zy-3 Tri-Stereo Imagery

Contact Info

Product

Resources

About