research-article

Accelerating MATLAB Image Processing Toolbox functions on GPUs

Authors:

Yi Yang,

Lin Cao,

Huiyang ZhouAuthors Info & Claims

GPGPU-3: Proceedings of the 3rd Workshop on General-Purpose Computation on Graphics Processing Units

Pages 75 - 85

https://doi.org/10.1145/1735688.1735703

Published: 14 March 2010 Publication History

Get Access

Abstract

In this paper, we present our effort in developing an open-source GPU (graphics processing units) code library for the MATLAB Image Processing Toolbox (IPT). We ported a dozen of representative functions from IPT and based on their inherent characteristics, we grouped these functions into four categories: data independent, data sharing, algorithm dependent and data dependent. For each category, we present a detailed case study, which reveals interesting insights on how to efficiently optimize the code for GPUs and highlight performance-critical hardware features, some of which have not been well explored in existing literature. Our results show drastic speedups for the functions in the data-independent or data-sharing category by leveraging hardware support judiciously; and moderate speedups for those in the algorithm-dependent category by careful algorithm selection and parallelization. For the functions in the last category, fine-grain synchronization and data-dependency requirements are the main obstacles to an efficient implementation on GPUs.

References

[1]

H. Breu et. al., Linear Time Euclidean Distance Transform Algorithms, The Insight Journal, 2006.

Google Scholar

[2]

N. Cuntz and Andreas Kolb, Fast Hierarchical 3D Distance Transforms on the GPU, EUROGRAPHICS 2007.

Google Scholar

[3]

P. Danielson, Euclidean distance mapping", Comput. Graph. Image Proc. 14, 227--248., 1980.

Crossref

Google Scholar

[4]

Y. Dotsenko et. al, Fast Scan Algorithms on Graphics Processors, ACM ICS 2008.

Digital Library

Google Scholar

[5]

R. Fabbri, et al., 2D Euclidean Distance Transform Algorithms: A Comparative Survey, ACM Computing Surveys, 2008.

Digital Library

Google Scholar

[6]

R. Floyd, L. Steinberg, An adaptive algorithm for spatial grayscale, Proceeding of the Society for Information Display, 1976.

Google Scholar

[7]

A. Meijster, et al, A general algorithm for computing distance transforms in linear time, ISMM, 2000.

Google Scholar

[8]

P. Metaxas, Parallel digital halftoning by error-diffusion, ACM International Conference Proceeding Series, 1995.

Google Scholar

[9]

J. D. Owens, M. Houston, D. Luebke, S. Green, J. E. Stone, J. C. Phillips. GPU Computing. Proceedings of the IEEE, 2008.

Google Scholar

[10]

G. Rong, T. Tan, Jump flooding in gpu with applications to Vornoi diagram and distance transform, ACM I3D, 2006.

Digital Library

Google Scholar

[11]

T. Saito and J. Toriwaki, New algorithms for Euclidean distance transformations of an n-dimensional digitised picture with applications, Patt. Recog. 27, 11, 1551--1565., 1994.

Crossref

Google Scholar

[12]

J. Schneider et. al, GPU-Based Real-Time Discrete Euclidean Distance Transforms With Precise Error Bounds, VISAPP 2009.

Google Scholar

[13]

S. Sengupta, M. Harris, Y. Zhang, J. Owens, Scan Primitives for GPU Computing, Graphics Hardware, 2007.

Digital Library

Google Scholar

[14]

T. Stockham, High speed convolution and correlation with application to digital filtering, Digital Processing of Signals, McGraw-Hill, 1969.

Google Scholar

[15]

https://sites.google.com/site/iptatiproject/

Google Scholar

[16]

ATI Stream Computing User Guide, 2009.

Google Scholar

[17]

ATI Radeon#8482; HD 5870 GPU Feature Summary, http://www.amd.com/us/products/desktop/graphics/ati-radeon-hd-5000/hd-5870/Pages/ati-radeon-hd-5870-specifications.aspx.

Google Scholar

[18]

Jacket: a GPU engine for MATLAB, http://www.accelereyes.com, 2009.

Google Scholar

[19]

MATLAB Acceleration. http://www.nvidia.com/object/matlab_acceleration.html

Google Scholar

[20]

MATLAB MEX files guide, http://www.mathworks.com/support/tech-notes/1600/1605.html, 2009.

Google Scholar

[21]

MATLAB plug-in for CUDA, http://developer.NVIDIA.com/object/MATLAB_cuda.html, 2009.

Google Scholar

[22]

NVIDIA CUDA Programming Guide, Version 2.2, 2009

Google Scholar

[23]

OpenCL - The open standard for parallel programming of heterogeneous systems, http://www.khronos.org/opencl/, 2009.

Google Scholar

Cited By

View all

Bartczak MPilarek M(2023)The Colourimetric Method for Mixing Time Measurement in Single-Use and Multi-Use Bioreactors—Methodology Overview and Practical RecommendationsEnergies10.3390/en1701022117:1(221)Online publication date: 31-Dec-2023
https://doi.org/10.3390/en17010221
Anand RMishra RKhan R(2022)Plant diseases detection using artificial intelligenceApplication of Machine Learning in Agriculture10.1016/B978-0-323-90550-3.00007-2(173-190)Online publication date: 2022
https://doi.org/10.1016/B978-0-323-90550-3.00007-2
Fu XJiang D(2022)High-throughput phenotyping: the latest research tool for sustainable crop production under global climate change scenariosSustainable Crop Productivity and Quality Under Climate Change10.1016/B978-0-323-85449-8.00003-8(313-381)Online publication date: 2022
https://doi.org/10.1016/B978-0-323-85449-8.00003-8
Show More Cited By

Index Terms

Accelerating MATLAB Image Processing Toolbox functions on GPUs
1. Computing methodologies
  1. Concurrent computing methodologies
    1. Concurrent programming languages
2. Software and its engineering
  1. Software notations and tools
    1. General programming languages
      1. Language types
        Concurrent programming languages

Recommendations

Accelerating financial applications on the GPU
GPGPU-6: Proceedings of the 6th Workshop on General Purpose Processor Using Graphics Processing Units

The QuantLib library is a popular library used for many areas of computational finance. In this work, the parallel processing power of the GPU is used to accelerate QuantLib financial applications. Black-Scholes, Monte-Carlo, Bonds, and Repo code paths ...
Accelerating a hydrological uncertainty ensemble model using graphics processing units (GPUs)

The practical application of hydrological uncertainty models that are designed to generate multiple ensembles can be severely restricted by the available computer processing power and thus, the time taken to generate the results. CPU clusters can help ...
A Portable and High-Performance General Matrix-Multiply (GEMM) Library for GPUs and Single-Chip CPU/GPU Systems
PDP '14: Proceedings of the 2014 22nd Euromicro International Conference on Parallel, Distributed, and Network-Based Processing

OpenCL is a vendor neutral and portable interface for programming parallel compute devices such as GPUs. Tuning OpenCL implementations of important library functions such as dense general matrix multiply (GEMM) for a particular device is a difficult ...

Reviews

Reviewer: Sergei Gorlatch

Graphics processing units (GPUs) are powerful general-purpose accelerators that can make virtually every personal computer (PC) run as a parallel computer. Many projects try to exploit the remarkable computational power of GPUs for data-parallel computations in MATLAB. While these projects usually address MATLAB as a whole, Kong et al. claim that they have done some of the first work on parallelizing MATLAB functions for image processing. The authors describe their efforts on porting 12 functions for image processing to GPUs produced by NVIDIA and AMD. The studied functions are a representative subset of functions provided by the MATLAB Image Processing Toolbox (IPT). The authors introduce four algorithm classes, in order to differentiate the functions according to their possible degree of parallelization. For each algorithm class, they present detailed optimization strategies for GPU-based implementations of functions of the respective class, as well as experimental performance results. A major contribution of this paper is its consistent comparison of implementations for AMD and NVIDIA GPUs using open computing language (OpenCL). Many previous papers on GPU programming were restricted to a single architecture or programming model. The paper also describes hardware-specific challenges and implementation strategies, making it a useful source of practical advice for developers of GPU-based applications. The authors successfully create a set of high-quality GPU-based image processing functions. Depending on the algorithm class, the experimental results show promising speedup values when compared with non-GPU implementations. Online Computing Reviews Service

Access critical reviews of Computing literature here

Become a reviewer for Computing Reviews.

Comments

Please enable JavaScript to view thecomments powered by Disqus.

Information & Contributors

Information

Published In

GPGPU-3: Proceedings of the 3rd Workshop on General-Purpose Computation on Graphics Processing Units

March 2010

124 pages

ISBN:9781605589350

DOI:10.1145/1735688

General Chairs:
David Kaeli
Northeastern University, Boston, MA
,
Miriam Leeser
Northeastern University, Boston, MA

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected]

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 14 March 2010

Permissions

Request permissions for this article.

Request Permissions

Check for updates

Author Tags

Qualifiers

Research-article

Funding Sources

Division of Computing and Communication Foundations

Conference

GPGPU-3

GPGPU-3: Third Workshop on General-Purpose Computation on Graphics Processing Units

March 14, 2010

Pennsylvania, Pittsburgh, USA

Acceptance Rates

Overall Acceptance Rate 57 of 129 submissions, 44%

Contributors

Other Metrics

View Article Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

26
Total Citations
View Citations
993
Total Downloads

Downloads (Last 12 months)12
Downloads (Last 6 weeks)2

Reflects downloads up to 20 Sep 2024

Other Metrics

View Author Metrics

Citations

Cited By

View all

Bartczak MPilarek M(2023)The Colourimetric Method for Mixing Time Measurement in Single-Use and Multi-Use Bioreactors—Methodology Overview and Practical RecommendationsEnergies10.3390/en1701022117:1(221)Online publication date: 31-Dec-2023
https://doi.org/10.3390/en17010221
Anand RMishra RKhan R(2022)Plant diseases detection using artificial intelligenceApplication of Machine Learning in Agriculture10.1016/B978-0-323-90550-3.00007-2(173-190)Online publication date: 2022
https://doi.org/10.1016/B978-0-323-90550-3.00007-2
Fu XJiang D(2022)High-throughput phenotyping: the latest research tool for sustainable crop production under global climate change scenariosSustainable Crop Productivity and Quality Under Climate Change10.1016/B978-0-323-85449-8.00003-8(313-381)Online publication date: 2022
https://doi.org/10.1016/B978-0-323-85449-8.00003-8
Zhao YNiu FZhang ZLi XChen JYang J(2020)Signal Detection and Enhancement for Seismic Crosscorrelation Using the Wavelet-Domain Kalman FilterSurveys in Geophysics10.1007/s10712-020-09620-6Online publication date: 17-Oct-2020
https://doi.org/10.1007/s10712-020-09620-6
Ruiz de Miras J(2019)Fast differential box-counting algorithm on GPUThe Journal of Supercomputing10.1007/s11227-019-03030-1Online publication date: 17-Oct-2019
https://doi.org/10.1007/s11227-019-03030-1
SALADRA DKOPERNIK M(2016)Qualitative and quantitative interpretation of SEM image using digital image processingJournal of Microscopy10.1111/jmi.12431264:1(102-124)Online publication date: 15-Jun-2016
https://doi.org/10.1111/jmi.12431
Yanju Zhu Shuguo Xie (2016)GPU acceleration for phase retrieval for electromagnetic interference source image2016 Asia-Pacific International Symposium on Electromagnetic Compatibility (APEMC)10.1109/APEMC.2016.7523020(237-239)Online publication date: May-2016
https://doi.org/10.1109/APEMC.2016.7523020
Lingala SZhu YKim YToutios ANarayanan SNayak K(2016)A fast and flexible MRI system for the study of dynamic vocal tract shapingMagnetic Resonance in Medicine10.1002/mrm.2609077:1(112-125)Online publication date: 17-Jan-2016
https://doi.org/10.1002/mrm.26090
Naik VKusur C(2015)Analysis of performance enhancement on graphic processor based heterogeneous architecture: A CUDA and MATLAB experiment2015 National Conference on Parallel Computing Technologies (PARCOMPTECH)10.1109/PARCOMPTECH.2015.7084519(1-5)Online publication date: Feb-2015
https://doi.org/10.1109/PARCOMPTECH.2015.7084519
Karas PMorard VBartovský JGrandpierre TDokládalová EMatula PDokládal P(2015)GPU implementation of linear morphological openings with arbitrary angleJournal of Real-Time Image Processing10.1007/s11554-012-0248-710:1(27-41)Online publication date: 1-Mar-2015
https://dl.acm.org/doi/10.1007/s11554-012-0248-7
Show More Cited By

View Options

Get Access

Login options

Check if you have access through your login credentials or your institution to get full access on this article.

Cited By

Index Terms

Recommendations

Accelerating financial applications on the GPU

Accelerating a hydrological uncertainty ensemble model using graphics processing units (GPUs)

A Portable and High-Performance General Matrix-Multiply (GEMM) Library for GPUs and Single-Chip CPU/GPU Systems

Reviews

Access critical reviews of Computing literature here