Computer Science - Research and Development

, Volume 26, Issue 1–2, pp 97–105 | Cite as

Discriminative Generalized Hough transform for localization of joints in the lower extremities

  • Heike RuppertshofenEmail author
  • Cristian Lorenz
  • Sarah Schmidt
  • Peter Beyerlein
  • Zein Salah
  • Georg Rose
  • Hauke Schramm
Special Issue Paper


A fully automatic iterative training approach for the generation of discriminative shape models for usage in the Generalized Hough Transform (GHT) is presented. The method aims at capturing the shape variability of the target object contained in the training data as well as identifying confusable structures (anti-shapes) and integrating this information into one model. To distinguish shape and anti-shape points and to determine their importance, an individual positive or negative weight is estimated for each model point by means of a discriminative training technique. The model is built from edge points surrounding the target point and the most confusable structure as identified by the GHT. Through an iterative approach, the performance of the model is gradually improved by extending the training dataset with images, where the current model failed to localize the target point. The proposed method is successfully tested on a set of 670 long-leg radiographs, where it achieves a localization rate of 74–97% for the respective tasks.


Object localization Generalized Hough transform Discriminative training Optimal model generation 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Ballard DH (1981) Generalizing the Hough transform to detect arbitrary shapes. Pattern Recogn 13(2):111–122 zbMATHCrossRefGoogle Scholar
  2. 2.
    Beyerlein P (1998) Discriminative model combination. In: IEEE international conference on acoustics, speech and signal processing. IEEE Press, New York, pp 481–484 Google Scholar
  3. 3.
    Deselaers T, Keysers D, Ney H (2005) Improving a discriminative approach to object recognition using image patches. In: 27th annual symposium of the German association for pattern recognition. LNCS, vol 3663. Springer, Heidelberg, pp 326–333 Google Scholar
  4. 4.
    Gall J, Lempitsky V (2009) Class-specific Hough forests for object detection. In: IEEE conference on computer vision and pattern recognition. IEEE Press, New York Google Scholar
  5. 5.
    Gooßen A, Schlüter M, Pralow T, Grigat RR (2010) A stitching algorithm for automatic registration of digital radiographs. In: International conference on image analysis and recognition. LNCS, vol 5112. Springer, Heidelberg, pp 854–862 Google Scholar
  6. 6.
    Gooßen A, Hermann E, Gernoth T, Pralow T, Grigat RR (2010) Model-based lower limb segmentation using weighted multiple candidates. In: Bildverarbeitung für die Medizin. Springer, Berlin, pp 276–280 Google Scholar
  7. 7.
    Heimann T, Münziger S, Meinzer HP et al. (2007) A shape-guided deformable model with evolutionary algorithm initialization for 3D soft tissue segmentation. In: International conference on information processing in medical imaging, pp 1–12 CrossRefGoogle Scholar
  8. 8.
    Heimann T, van Ginneken B, Styner M et al. (2009) Comparison and evaluation of methods for liver segmentation from CT datasets. IEEE Trans Med Imaging 28(8):1251–1265 CrossRefGoogle Scholar
  9. 9.
    Jaynes ET (1957) Information theory and statistical mechanics. Phys Rev 106(4):620–630 CrossRefMathSciNetGoogle Scholar
  10. 10.
    Lee Y, Hara T, Fujita H, Itoh S, Ishigaki T (2001) Automated detection of pulmonary nodules in helical CT images based on an improved template-matching technique. IEEE Trans Med Imaging 20(7):595–604 CrossRefGoogle Scholar
  11. 11.
    Leibe B, Leonardis A, Schiele B (2008) Robust object detection with interleaved categorization and segmentation. Int J Comput Vis 77(1–3):259–289 CrossRefGoogle Scholar
  12. 12.
    Maji S, Malik J (2009) Object detection using a max-margin Hough transform. In: IEEE conference on computer vision and pattern recognition. IEEE Press, New York, pp 1038–1045 CrossRefGoogle Scholar
  13. 13.
    Recuero ABM, Beyerlein P, Schramm H (2008) Discriminative optimization of 3D shape models for the Generalized Hough transform. In: AMIES Kiel Google Scholar
  14. 14.
    Ruppertshofen H, Lorenz C, Beyerlein P, Salah Z, Rose G, Schramm H (2010) Fully automatic model creation for object localization utilizing the Generalized Hough transform. In: Bildverarbeitung für die Medizin. Springer, Berlin, pp 281–285 Google Scholar
  15. 15.
    Schramm H, Ecabert O, Peters J et al. (2006) Towards fully automatic object detection and segmentation. In: SPIE medical imaging 2006: image processing, p 614402 Google Scholar
  16. 16.
    Seghers D, Slagmolen P, Lambelin Y et al. (2007) Landmark based liver segmentation using local shape and local intensity models. In: MICCAI workshop on 3D segmentation in the clinic: a grand challenge. Springer, Berlin, pp 135–142 Google Scholar
  17. 17.
    Zheng Y, Georgescu B, Comaniciu D (2009) Marginal space learning for efficient detection of 2D/3D anatomical structures in medical images. In: 21st international conference on information processing in medical imaging. LNCS, vol 5636. Springer, Heidelberg, pp 411–422 CrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2010

Authors and Affiliations

  • Heike Ruppertshofen
    • 1
    • 4
    Email author
  • Cristian Lorenz
    • 2
  • Sarah Schmidt
    • 3
    • 4
  • Peter Beyerlein
    • 3
  • Zein Salah
    • 4
  • Georg Rose
    • 4
  • Hauke Schramm
    • 1
  1. 1.Institute of Applied Computer ScienceUniversity of Applied Sciences KielKielGermany
  2. 2.Department Digital ImagingPhilips Research EuropeHamburgGermany
  3. 3.Department of EngineeringUniversity of Applied Sciences WildauWildauGermany
  4. 4.Institute of Electronics, Signal Processing and Communication TechnologyOtto-von-Guericke-UniversityMagdeburgGermany

Personalised recommendations