Skip to main content
SpringerLink
Log in

Parallelizing Multimodal Background Modeling on a Low-Power Integrated GPU

  • Published:
Journal of Signal Processing Systems Aims and scope Submit manuscript

Abstract

Background modeling techniques for embedded computer vision applications must balance accuracy, speed, and power. Basic background modeling techniques run quickly, but their accuracy is not sufficient for computer vision problems involving dynamic background. In contrast, adaptive background modeling techniques are more robust, but run more slowly. Due to its high inherent fine-grain parallelism, robust adaptive background modeling has been implemented on GPUs with significant performance improvements over CPUs. However, these implementations are infeasible in embedded applications due to the high power ratings of the targeted general-purpose GPU platforms. This paper focuses on exploiting fine-grain data parallelism and optimizing memory access patterns to target a low-cost adaptive background modeling algorithm multimodal mean (MMM) to a low-power GPU with thermal design power (TDP) of only 12 watts. The algorithm has comparable accuracy with the Gaussian mixture model (GMM) algorithm, but less computational and memory cost. It achieves a frame rate of 392 fps with a full VGA resolution (640x480) frame on the low-power integrated GPU NVIDIA ION. This is a 20x speed-up of the MMM algorithm compared to the embedded CPU platform Intel Atom of comparable TDP. In addition, the MMM algorithm attains a 5-6x speed up over the GMM implementation on the ION GPU platform.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10

Similar content being viewed by others

References

  1. Test images for wallflower paper (1999). http://research.microsoft.com/en-us/um/people/jckrumm/wallflower/testimages.htm.

  2. PETS 2009 benchmark data (2009). http://www.cvg.reading.ac.uk/PETS2009/a.html#s111.

  3. Intellio ILC-BL series smart cameras. http://www.videoline-tvcc.com/upload/pdf/ILC-BL_series_datasheet_ENG.pdf. accessed November 2015.

  4. Matrox IRIS-GT smart camera. http://www.matrox.com/imaging/media/pdf/products/iris_gt_da/iris_gt_da.pdf. accessed November 2015.

  5. Sony XCISX100C-XP smart camera. http://pro.sony.com/bbsc/ssr/cat-camerasindustrial/cat-cismartcameras/product-XCISX100C%2FXP/. accessed November 2015.

  6. NI-177x series smart cameras. http://sine.ni.com/ds/app/doc/p/id/ds-370/lang/en. accessed November 2015.

  7. Apewokin, S., Valentine, B., Forsthoefel, D., Wills, L., Wills, S., & Gentile, A. (2010). Embedded real-time surveillance using multimodal mean background modeling. In kisacanin, B., Bhattacharyya, S., & Chai, S. (Eds.) Embedded computer vision, (pp. 163–175.): Springer.

  8. Azmat, S., Wills, L., & Wills, S. (2012). Accelerating adaptive background modeling on low-power integrated GPUs. In International workshop on embedded multicore systems (ICPP-EMS 2012), held in conjunction with the 41st IEEE international conference on parallel processing (pp. 568–573).

  9. Carr, P. (2008). GPU accelerated multimodal background subtraction. In Digital image computing: Techniques and applications (DICTA) (pp. 279–286).

  10. Fabiàn, T., & Gaura, J. (2008). Parallel implementation of recursive background modeling technique in CUDA for tracking moving objects in video traffic surveillance. In 4th Doctoral Workshop on Mathematical and Engineering Methods in Computer Science. http://www.fi.muni.cz/memics07/2008/pres/fabian_cuda.pdf.

  11. Horprasert, T., Harwood, D., & Davis, L.S. (1999). A statistical approach for real-time robust background subtraction and shadow detection. In IEEE International conferecne on computer vision (ICCV), (Vol. 99 pp. 1–19).

  12. Hsieh, K.Y., Lai, C.H., Lai, S.H., & Lee, J.K. (2012). Parallelization of belief propagation on cell processors for stereo vision. ACM Transactions on Embedded Computing Systems (TECS), 11(1), 13.

    Google Scholar 

  13. Kirk, D.B., & Wen-mei, W.H. (2012). Programming massively parallel processors: a hands-on approach Newnes.

  14. Liu, Y., & Hu, J. (2011). GPU-based parallelization for fast circuit optimization. ACM Transactions on Design Automation of Electronic Systems (TODAES), 16(3), 24.

    Article  Google Scholar 

  15. NVIDIA Corporation: NVIDIA Compute Unified Device Architecture C Programming Guide v6.5. http://docs.nvidia.com/cuda/pdf/CUDA_C_Programming_Guide.pdf. accessed January 2015.

  16. NVIDIA Corporation: NVIDIA Compute Unified Device Architecture C Best Practices Guide v6.5. http://docs.nvidia.com/cuda/pdf/CUDA_C_Best_Practices_Guide.pdf. accessed January 2015.

  17. Pham, V., Vo, P., & Hung, V.T. (2010). GPU implementation of extended gaussian mixture model for background subtraction. In 2010 IEEE RIVF International conference on computing and communication technologies, research, innovation, and vision for the future (pp. 1–4).

  18. Poremba, M., Xie, Y., & Wolf, M. (2010). Accelerating adaptive background subtraction with GPU and CBEA architecture. In 2010 IEEE Workshop on signal processing systems (SIPS) (pp. 305–310).

  19. Scogland, T.R.W., Lin, H., & Feng, W. (2010). A first look at integrated GPUs for green high-performance computing. Computer Science - Research and Development, 25(3–4), 125–134.

    Article  Google Scholar 

  20. Sen-Ching, S.C., & Kamath, C. (2004). Robust techniques for background subtraction in urban traffic video. In Electronic imaging 2004. International society for optics and photonics (pp. 881–892).

  21. Stauffer, C., & Grimson, W.E.L. (2000). Learning patterns of activity using real-time tracking. In IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), (Vol. 22 pp. 747–757).

  22. Volkov, V. (2010). Better performance at lower occupancy (Presentation in GPU Technology Conference. http://www.cs.berkeley.edu/~volkov/volkov10-GTC.pdf.

  23. Williams, S., Shalf, J., Oliker, L., Kamil, S., Husbands, P., & Yelick, K. (2006). The potential of the cell processor for scientific computing. In Proceedings of the 3rd conference on computing frontiers, ACM (pp. 9–20).

  24. Zhu, Y., Wang, B., & Deng, Y. (2011). Massively parallel logic simulation with GPUs. ACM Transactions on Design Automation of Electronic Systems (TODAES), 16(3), 29.

    Article  Google Scholar 

  25. Zivkovic, Z., & van der Heijden, F. (2006). Efficient adaptive density estimation per image pixel for the task of background subtraction. Pattern Recognition Letters, 27(7), 773–780.

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Electrical Engineering, Comsats Institute of Information Technology, Abbottabad, Pakistan

    Shoaib Azmat

  2. School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, USA

    Linda Wills & Scott Wills

Authors
  1. Shoaib Azmat
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Linda Wills
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Scott Wills
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Shoaib Azmat.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Azmat, S., Wills, L. & Wills, S. Parallelizing Multimodal Background Modeling on a Low-Power Integrated GPU. J Sign Process Syst 88, 43–53 (2017). https://doi.org/10.1007/s11265-016-1111-z

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11265-016-1111-z

Keywords

Search

Navigation