Advertisement

A Scalable Dataflow Accelerator for Real Time Onboard Hyperspectral Image Classification

  • Shaojun Wang
  • Xinyu Niu
  • Ning Ma
  • Wayne Luk
  • Philip Leong
  • Yu Peng
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9625)

Abstract

Real-time hyperspectral image classification is a necessary primitive in many remotely sensed image analysis applications. Previous work has shown that Support Vector Machines (SVMs) can achieve high classification accuracy, but unfortunately it is very computationally expensive. This paper presents a scalable dataflow accelerator on FPGA for real-time SVM classification of hyperspectral images.To address data dependencies, we adapt multi-class classifier based on Hamming distance. The architecture is scalable to high problem dimensionality and available hardware resources. Implementation results show that the FPGA design achieves speedups of 26x, 1335x, 66x and 14x compared with implementations on ZYNQ, ARM, DSP and Xeon processors. Moreover, one to two orders of magnitude reduction in power consumption is achieved for the AVRIS hyperspectral image datasets.

References

  1. 1.
    Bioucas-Dias, J.M., et al.: Hyperspectral remote sensing data analysis and future challenges. IEEE Geosci. Remote Sens. Mag. 6, 6–36 (2013)CrossRefGoogle Scholar
  2. 2.
    Cadambi, S., Igor, D., et al.: A massively parallel FPGA-based coprocessor for support vector machines. In: Proceedings - IEEE Symposium on Field Programmable Custom Computing Machines, FCCM 2009, pp. 115–122 (2009)Google Scholar
  3. 3.
    Gustavo, C., Davis, T., et al.: Advances in hyperspectral image classification: earth monitoring with statistical learning methods. IEEE Sig. Process. Mag. 31(1), 45–54 (2014)CrossRefGoogle Scholar
  4. 4.
    Irick, K.M., et al.: A hardware efficient support vector machine architecture for FPGA. In: Proceedings of the 16th IEEE Symposium on Field-Programmable Custom Computing Machines, FCCM 2008, pp. 304–305 (2008)Google Scholar
  5. 5.
    Khodadadzadeh, M., et al.: A new framework for hyperspectral image classification using multiple spectral and spatial features. In: IEEE Geoscience and Remote Sensing Symposium, pp. 4628–4631 (2014)Google Scholar
  6. 6.
    Kyrkou, C., Theocharides, T.: SCoPE: towards a systolic array for SVM object detection. IEEE Embed. Syst. Lett. 1(2), 46–49 (2009)CrossRefGoogle Scholar
  7. 7.
    Liu, Y., et al.: Hyperspectral classification via deep networks and superpixel segmentation. Int. J. Remote Sens. 36(13), 3459–3482 (2015)CrossRefGoogle Scholar
  8. 8.
    Lopez, S., et al.: The promise of reconfigurable computing for hyperspectral imaging onboard systems: a review and trends. Proc. IEEE 101(3), 698–722 (2013)CrossRefGoogle Scholar
  9. 9.
    Melgani, F., Bruzzone, L.: Classification of hyperspectral remote sensing images with support vector machines. IEEE Trans. Geosci. Remote Sens. 42(8), 1778–1790 (2004)CrossRefGoogle Scholar
  10. 10.
    Montenegro, S., et al.: Hyperspectral monitoring data processing, pp. 1–4 (2003). ISBN 3-89685-569-7Google Scholar
  11. 11.
    Papadonikolakis, M., Bouganis, C.S.: Novel cascade FPGA accelerator for support vector machines classification. IEEE Trans. Neural Netw. Learn. Syst. 23(7), 1040–1052 (2012)CrossRefGoogle Scholar
  12. 12.
    Papadonikolakis, M., Bouganis, C.S.: A heterogeneous FPGA architecture for support vector machine training. In: 18th IEEE Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM), pp. 6–9 (2010)Google Scholar
  13. 13.
    Sami, Q., et al.: Neural network based adaboosting approach for hyperspectral data classification. In: International Conference on Computer Science and Network Technolgoy, pp. 241–245 (2011)Google Scholar
  14. 14.
    Christos, K., et al.: Embedded hardware-efficient real-time classification with cascade support vector machines. IEEE Trans. Neural Netw. Learn. Syst. 26(1), 99–112 (2016)Google Scholar
  15. 15.
    Xue, Z., et al.: Harmonic analysis for hyperspectral image classification integrated with PSO optimized SVM. IEEE J. Sel. Top. Appl. Earth Observations Remote Sens. 7(6), 2131–2146 (2014)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  • Shaojun Wang
    • 1
    • 2
  • Xinyu Niu
    • 2
  • Ning Ma
    • 1
  • Wayne Luk
    • 2
  • Philip Leong
    • 3
  • Yu Peng
    • 1
  1. 1.Harbin Institute of TechnologyHarbinChina
  2. 2.Imperial College LondonLondonUK
  3. 3.University of SydneySydneyAustralia

Personalised recommendations