An Improved SIFT Algorithm Based on Invariant Gradient

Conference paper
First Online:
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 369)

Abstract

In order to make the feature descriptor stable for rotating, the SIFT (Scale Invariant Feature Transform) algorithm assigned a main direction for feature points and rotated the local image according to the main direction. This paper do some research on the rotating process of SIFT algorithm, and put forward a new algorithm based on invariant gradient. The defined pixels’ gradient-invariant in the new algorithm is mainly relevant to the gray value of the nearest 8 pixels, and has nothing with the relative position of the 8 pixels around. The experimental results showed that collecting pixels’ gradient-invariant statistics can effectively improve SIFT algorithm’s computing speed.

Keywords

Invariant gradient Feature detection SIFT 
This is a preview of subscription content, log in to check access.

References

  1. 1.
    Lowe, D. G. (2004). Distinctive image features from scale invariant keypoints. International Journal of Computer Vision, 2(60), 91–110.CrossRefGoogle Scholar
  2. 2.
    Lowe, D. G. (1999). Object recognition from local scale- invariant features. In Proceedings of the 7th IEEE International Conference on Computer Vision, Kerkyra (pp. 1150–1157). Greece: IEEE.Google Scholar
  3. 3.
    Zheng, Y. B., Huang, X. S., & Feng, S. J. (2010). An image matching algorithm combining SIFT with LBP which is invariant for rotation. Journal of Computer Aided Design and Graphics, 22(2), 287–292.Google Scholar
  4. 4.
    Tang, C. W., & Xiao, J. (2012). An improved SIFT descriptor with analysis. Journal of Wuhan University, 37(1), 11–16.Google Scholar
  5. 5.
    Yang, K., & Sukthankar, R. (2004). PCA-SIFT: A more distinctive representation for local image descriptors. In The IEEE Conference on Computer Visionand Pattern Recognition, Washington, DC, USA.Google Scholar
  6. 6.
    Wan, X., Zhang, Z. X., & Ke, T. (2013). An improved SIFT algorithm based on the zero crossing theory. Journal of Wuhan University, 38(3), 270–273.Google Scholar
  7. 7.
    Mo, H. Y., & Wang, Z. P. (2011). A feature detection algorithm combining MSER and SIFT. Journal of Dongbei University, 37(5), 624–628.Google Scholar
  8. 8.
    Wang, S. (2013). SIFT based image matching algorithm research. MS Thesis, Xi’an Electronic technology University.Google Scholar
  9. 9.
    Feng, J. (2010). The research and improvement of SIFT. MS Thesis, Jilin University.Google Scholar

Copyright information

© Springer Science+Business Media Singapore 2016

Authors and Affiliations

  1. 1.Zhengzhou Institute of Surveying and MappingZhengzhouChina

Personalised recommendations