Enhanced Scalar-Invariant Feature Transformation

  • S. Adithya
  • M. Sivagami
Conference paper
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 490)


The proposed work enhances the feature point detection in the scalar-invariant feature transformation (SIFT). The sequence of steps in the SIFT algorithm drops most of the feature points in the low-contrast regions of the image. This paper provides a solution to this problem by adding the output of Sobel filtered image on the input image iteratively until the entropy of the input image is increased to a saturation level. The SIFT descriptors generated for this enhanced image tend to describe the redundant features in the image. To overcome this problem, affinity propagation clustering is done.


Scalar-invariant feature transformation (SIFT) Low contrast Entropy Sobel Affinity propagation clustering 


  1. 1.
    Lowe DG (1999) Object recognition from local scale-invariant features. Proc Int Conf Comput Vis 2:1150–1157. doi:
  2. 2.
    Andono PN, Purnama I, Eddy K, Hariadi M (2013) Underwater image enhancement using adaptive filtering for enhanced SIFT-BASED image matching. J Theor Appl Inform Technol 52(3):273–280Google Scholar
  3. 3.
    Liangping Tu, Dong C (2013) In: 2013 6th international congress on image and signal processing (CISP), December 2013 pp 16–18Google Scholar
  4. 4.
    Castillo-Carrión S, Guerrero-Ginel J-E (2017) SIFT optimization and automation for matching images from multiple temporal sources. Int J Appl Earth Obs Geoinf 57:113–127Google Scholar
  5. 5.
    Guohua Lv, Teng SW, Lu G (2016) Enhancing SIFT-based image registration performance by building and selecting highly discriminating descriptors. Pattern Recogn Lett.
  6. 6.
    Witkin AP (1983) Scale-space filtering. In: Proceedings 8th international joint conference on artificial intelligence, Karlsruhe, Germany, pp 1019–1022Google Scholar
  7. 7.
    Dueck D (2009) Affinity propagation: clustering data by passing messages. University of TorontoGoogle Scholar
  8. 8.
    Li Q (2006) Research on handmetric recognition and feature level fusion method (Doctoral dissertation, PhD thesis, BeiJing JiaoTong University, Beijing)Google Scholar
  9. 9.
    Polesel A, Ramponi G, Mathews VJ (2000) Image enhancement via adaptive unsharp masking. IEEE Trans Image Process 9(3):505–510CrossRefGoogle Scholar
  10. 10.
    Mitra SK, Li H, Lin I, Yu T (1991) A new class of nonlinear filters for image enhancement. Acoust Speech Signal Process 4:2525–2528Google Scholar
  11. 11.
    Pizer SM, Amburn EP, Austin JD, Cromartie R, Geselowitz A, Greer T, ter Haar Romeny B, Zimmerman JB, Zuiderveld K (1987) Adaptive histogram equalization and its variations. Comput vision Graph image process 39(3):355–368Google Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2018

Authors and Affiliations

  1. 1.School of Computing Sciences and EngineeringVIT UniversityChennaiIndia

Personalised recommendations