Feature Detection and Description using a Harris-Hessian/FREAK Combination on an Embedded GPU

Max Danielsson, Thomas Sievert, Håkan Grahn, Jim Rasmusson

Abstract

GPUs in embedded platforms are reaching performance levels comparable to desktop hardware, thus it becomes interesting to apply Computer Vision techniques. We propose, implement, and evaluate a novel feature detector and descriptor combination, i.e., we combine the Harris-Hessian detector with the FREAK binary descriptor. The implementation is done in OpenCL, and we evaluate the execution time and classification performance. We compare our approach with two other methods, FAST/BRISK and ORB. Performance data is presented for the mobile device Xperia Z3 and the desktop Nvidia GTX 660. Our results indicate that the execution times on the Xperia Z3 are insufficient for real-time applications while desktop execution shows future potential. Classification performance of Harris-Hessian/FREAK indicates that the solution is sensitive to rotation, but superior in scale variant images.

References

  1. Alahi, A., Ortiz, R., and Vandergheynst, P. (2012). FREAK: Fast Retina Keypoint. In 2012 IEEE Conf. on Computer Vision and Pattern Recognition, pages 510-517.
  2. Bay, H., Tuytelaars, T., and Gool, L. V. (2006). SURF: Speeded Up Robust Features. In Computer Vision - ECCV 2006, number 3951 in LNCS, pages 404-417.
  3. Bekele, D., Teutsch, M., and Schuchert, T. (2013). Evaluation of binary keypoint descriptors. In 20th IEEE Int'l Conf. on Image Processing (ICIP), pages 3652-3656.
  4. Calonder, M., Lepetit, V., Strecha, C., and Fua, P. (2010). BRIEF: Binary Robust Independent Elementary Features. In Computer Vision - ECCV 2010 , number 6314 in LNCS, pages 778-792.
  5. Deriche, R. (1993). Recursively implementating the Gaussian and its derivatives. report.
  6. Harris, C. and Stephens, M. (1988). A Combined Corner and Edge Detector. volume 15, page 50, Manchester. The Plessey Company plc.
  7. Leutenegger, S., Chli, M., and Siegwart, R. (2011). BRISK: Binary Robust invariant scalable keypoints. In IEEE Int'l Conf. on Computer Vision, pages 2548-2555.
  8. Lowe, D. (1999). Object recognition from local scaleinvariant features. In Seventh IEEE Int'l Conf. on Computer Vision, 1999, pages 1150-1157 vol.2.
  9. Mikolajczyk, K. and Schmid, C. (2004). Scale & affine invariant interest point detectors. International Journal of Computer Vision, 60(1):63-86.
  10. Munshi, A. (2011). The OpenCL Specification Version: 1.1 Document Revision: 44.
  11. Rublee, E., Rabaud, V., Konolige, K., and Bradski, G. (2011). ORB: An efficient alternative to SIFT or SURF. In IEEE Int'l Conf. on Computer Vision, pages 2564-2571.
  12. Timm, C., Gelenberg, A., Weichert, F., and Marwedel, P. (2010). Reducing the energy consumption of embedded systems by integrating general purpose GPUs. Technical Report TR829, TU Dortmund, Dept. of Computer Science.
  13. Xie, H., Gao, K., Zhang, Y., Li, J., and Liu, Y. (2010). GPUbased fast scale invariant interest point detector. In 2010 IEEE Int'l Conf. on Acoustics Speech and Signal Processing (ICASSP), pages 2494-2497.
  14. Xie, H., Gao, K., Zhang, Y., Tang, S., Li, J., and Liu, Y. (2011). Efficient Feature Detection and Effective Post-Verification for Large Scale Near-Duplicate Image Search. IEEE Transactions on Multimedia, 13(6):1319-1332.
Download


Paper Citation


in Harvard Style

Danielsson M., Sievert T., Grahn H. and Rasmusson J. (2016). Feature Detection and Description using a Harris-Hessian/FREAK Combination on an Embedded GPU . In Proceedings of the 5th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-173-1, pages 517-525. DOI: 10.5220/0005662005170525


in Bibtex Style

@conference{icpram16,
author={Max Danielsson and Thomas Sievert and Håkan Grahn and Jim Rasmusson},
title={Feature Detection and Description using a Harris-Hessian/FREAK Combination on an Embedded GPU},
booktitle={Proceedings of the 5th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2016},
pages={517-525},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0005662005170525},
isbn={978-989-758-173-1},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 5th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - Feature Detection and Description using a Harris-Hessian/FREAK Combination on an Embedded GPU
SN - 978-989-758-173-1
AU - Danielsson M.
AU - Sievert T.
AU - Grahn H.
AU - Rasmusson J.
PY - 2016
SP - 517
EP - 525
DO - 10.5220/0005662005170525