Contour Context Selection for Object Detection: A Set-to-Set Contour Matching Approach

  • Qihui Zhu
  • Liming Wang
  • Yang Wu
  • Jianbo Shi
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5303)

Abstract

We introduce a shape detection framework called Contour Context Selection for detecting objects in cluttered images using only one exemplar. Shape based detection is invariant to changes of object appearance, and can reason with geometrical abstraction of the object. Our approach uses salient contours as integral tokens for shape matching. We seek a maximal, holistic matching of shapes, which checks shape features from a large spatial extent, as well as long-range contextual relationships among object parts. This amounts to finding the correct figure/ground contour labeling, and optimal correspondences between control points on/around contours. This removes accidental alignments and does not hallucinate objects in background clutter, without negative training examples. We formulate this task as a set-to-set contour matching problem. Naive methods would require searching over ’exponentially’ many figure/ground contour labelings. We simplify this task by encoding the shape descriptor algebraically in a linear form of contour figure/ground variables. This allows us to use the reliable optimization technique of Linear Programming. We demonstrate our approach on the challenging task of detecting bottles, swans and other objects in cluttered images.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Zhu, Q., Song, G., Shi, J.: Untangling cycles for contour grouping. In: ICCV (2007)Google Scholar
  2. 2.
    Belongie, S., Malik, J., Puzicha, J.: Shape matching and object recognition using shape contexts. PAMI (2002)Google Scholar
  3. 3.
    Wang, L., Shi, J., Song, G., Shen, I.F.: Object detection combining recognition and segmentation. In: ACCV (2007)Google Scholar
  4. 4.
    Ferrari, V., Fevrier, L., Jurie, F., Schmid, C.: Groups of adjacent contour segments for object detection. PAMI (2007)Google Scholar
  5. 5.
    Ferrari, V., Jurie, F., Schmid, C.: Accurate object detection with deformable shape models learnt from images. In: CVPR (2007)Google Scholar
  6. 6.
    Shotton, J., Blake, A., Cipolla, R.: Contour-based learning for object detection. In: ICCV (2005)Google Scholar
  7. 7.
    Amir, A., Lindenbaum, M.: Grouping-based nonadditive verification. PAMI (February 1998)Google Scholar
  8. 8.
    Felzenszwalb, P.F., Schwartz, J.D.: Hierarchical matching of deformable shapes. In: CVPR (2007)Google Scholar
  9. 9.
    Zahn, C.T., Roskies, R.S.: Fourier descriptors for plane closed curves. IEEE Trans. of Computing (March 1972)Google Scholar
  10. 10.
    Gdalyahu, Y., Weinshall, D.: Flexible syntactic matching of curves and its application to automatic hierarchical classification of silhouettes. PAMI (1999)Google Scholar
  11. 11.
    Lamdan, Y., Schwartz, J.T., Wolfson, H.J.: Affine invariant model-based object recognition. IEEE Trans. Robotics and Automation 6, 578–589 (1990)CrossRefGoogle Scholar
  12. 12.
    Amit, Y., Wilder, K.: Joint induction of shape features and tree classifiers. PAMI (1997)Google Scholar
  13. 13.
    Cootes, T.F., Taylor, C.J., Cooper, D.H., Graham, J.: Active shape models-their training and application. Computer Vision and Image Understanding 61(1), 38–59 (1995)CrossRefGoogle Scholar
  14. 14.
    Ling, H., Jacobs, D.W.: Using the inner-distance for classification of articulated shapes. In: ICCV (2005)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Qihui Zhu
    • 1
  • Liming Wang
    • 2
  • Yang Wu
    • 3
  • Jianbo Shi
    • 1
  1. 1.Department of Computer and Information ScienceUniversity of Pennsylvania 
  2. 2.Department of Computer Science and EngineeringFudan University 
  3. 3.Institue of Artificial Intelligence and RoboticsXi’an Jiaotong University 

Personalised recommendations