Advertisement

Large Scale Specific Object Recognition by Using GIFTS Image Feature

  • Hiroki NakanoEmail author
  • Yumi Mori
  • Chiaki Morita
  • Shingo Nagai
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9280)

Abstract

We propose GIFTS (Goods Image Features for Tree Search) which uses image local features for large-scale object recognition. Each GIFTS is a kind of keypoint feature. The feature vector consists of intensity deltas for 128 selected pixel pairs around the keypoint. By generating a KD-Tree from the GIFTS feature vectors of the training images and using the KD-Tree to search for nearest neighbor feature vectors of a query image, query times are on the order of log N for specific object recognition. We used the proposed method for book cover queries with 100,000 training images, had recognition accuracy over 99% with query times within one second.

Keywords

Object recognition Local feature KD-Tree Augmented reality 

References

  1. 1.
    Lowe, D.: Distinctive Image Features from Scale-Invariant Keypoints. International Journal of Computer Vision 60(2), 91–110 (2004)CrossRefGoogle Scholar
  2. 2.
    Ke, Y., Sukthankar, R.: PCA-SIFT: a more distinctive representation for local image descriptors. In: CVPR, vol. 2, pp. 506–513 (2004Google Scholar
  3. 3.
    Bay, H., Tuytelaars, T., Van Gool, L.: Surf: speeded up robust features. In: Leonardis, A., Bischof, H., Pinz, A. (eds.) ECCV 2006, Part I. LNCS, vol. 3951, pp. 404–417. Springer, Heidelberg (2006)CrossRefGoogle Scholar
  4. 4.
    Juan, L., Gwun, O.: A Comparison of SIFTPCA-SIFT and SURF. International Journal of Image Processing 3(4), 143–152 (2009)Google Scholar
  5. 5.
    Rublee, E., et. al.: ORB: an efficient alternative to SIFT or SURF. In: ICCV2011, pp. 2564–2571 (2011)Google Scholar
  6. 6.
    Arya, S., et al.: An Optimal Algorithm for Approximate Nearest Neighbor Searching in Fixed Dimensions. Journal of the ACM 45(6), 891–923 (1998)MathSciNetCrossRefzbMATHGoogle Scholar
  7. 7.
    Andoni, A.: Near-optimal hashing algorithms for approximate nearest neighbor in high dimensions. In: 47th Annual IEEE Symposium on Foundations of Computer Science, pp. 459–468 (2006)Google Scholar
  8. 8.
    Muja, M., Lowe, D.: Fast matching of binary features. In: Conference On Computer And Robot Vision (CRV 2012), pp. 404–410 (2012)Google Scholar
  9. 9.
    Kise, K., Noguchi, K., Iwamura, M.: Simple representation and approximate search of feature vectors for large-scale object recognition. In: Proceedings of BMVC 2007, vol. 1, pp. 182–191 (2007)Google Scholar
  10. 10.
    Muja, M., Lowe, D.: FLANN: Fast library for approximate nearest neighbors. http://www.cs.ubc.ca/research/flann
  11. 11.
    Bekel, D., Teutschy, M., Schucherty, T.: Evaluation of binary keypoint descriptors. In: ICIP 2013, pp. 3652–3656 (2013)Google Scholar
  12. 12.
    Do, A.T., Ilango, K., Ramasamy, D., Kalidasan, S., Balakrinan, V., Chang, R.T.: Effectiveness of low vision services in improving patient quality of life at Aravind Eye Hospital. Indian J. Ophthalmol 62(12), 1125–1131 (2014)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  • Hiroki Nakano
    • 1
    Email author
  • Yumi Mori
    • 2
  • Chiaki Morita
    • 1
  • Shingo Nagai
    • 1
  1. 1.Tokyo Laboratory, IBM Japan Ltd.TokyoJapan
  2. 2.Yokohama National UniversityYokohamaJapan

Personalised recommendations