Skip to main content
Log in

Object tracking in image sequences based on parametric features

Objektverfolgung in Bildsequenzen auf der Basis parametrischer Merkmale

  • Reviewed Original Papers
  • Published:
e&i Elektrotechnik und Informationstechnik Aims and scope Submit manuscript

Abstract

In this paper a feature based algorithm for tracking moving objects with an active camera system is presented. It uses oriented structure elements like edges or lines for the estimation of motion-induced object displacements in gray-level input images. After an initial Gabor filtering process, a spatially extended structure consisting of simple local features in the filter response is merged into a vector of more complex features through a parameterization process. Corresponding vectors in two subsequent frames are detected by iteratively computing a similarity measure for all feature vectors. This enables the detection of larger object displacements because a proper parameterization leads to highly discriminable feature vectors describing certain image structures.

Zusammenfassung

In diesem Beitrag wird ein merkmalsbasierter Algorithmus für die Verfolgung bewegter Objekte durch ein aktives Sehsystem vorgestellt. Es werden orientierte Strukturelemente wie Kanten oder Linien benutzt, um die durch Bewegung verursachte Verschiebung eines Objekts in Grauwertbildern zu schätzen. Im Anschluss an einer Vorverarbeitung der Eingangsbilder mittels Gaborfiltern werden räumlich ausgedehnte Strukturen, bestehend aus einfachen lokalen Merkmalen in der Filterantwort, im Rahmen eines Parametrisierungsprozesses zu komplexeren Merkmalvektoren zusammengefasst. Korrespondierende Vektoren in zwei aufeinanderfolgenden Bildern werden durch die iterative Berechnung eines Ähnlichkeitsmaßes fur alle Merkmalvektoren erkannt. Dies ermöglicht die Detektion größerer Objektverschiebungen, da eine geeignete Parametrisierung zu gut unterscheidbaren Merkmalvektoren führt, welche die einzelnen Bildstrukturen beschreiben.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

References

  1. Andersson, M.: Tracking methods in computer vision. Ph. D. thesis, Computational Vision and Active Perception Laboratory, Royal Institute of Technology. Stockholm. 1994.

    Google Scholar 

  2. Bollmann, M., Hoischen, R., Mertsching, B.: Integration of static and dynamics scene features guiding visual attention. In: Paulus, E., Wahl, F. M. (eds): Mustererkennung 1997. Berlin: Springer. 1997, pp. 483–490.

    Google Scholar 

  3. Mertsching, B., Bollmann, M., Hoischen, R., Schmalz, S.: The Neural Active Vision System NAVIS. To appear in: Jähne, B., Haußecker, H., Geißler, P. (eds.): Handbook of computer vision and applications. San Diego: Academic Press. 1999.

    Google Scholar 

  4. Chang, Y.-L., Aggarwal, J. K.: Line correspondences from cooperating spatial and temporal grouping processes for a sequence of images. Computer Vision and Image Understanding 67 (1997), No. 2, pp. 186–201.

    Article  Google Scholar 

  5. Mitiche, A, Bouthemy, P: Computation and analysis of image motion: a synopsis of current problems and methods. International Journal of Computer Vision 19 (1996), No. 1, pp. 29–55.

    Article  Google Scholar 

  6. Reid, I, Murray, D: Active tracking of foveated feature clusters using affine structure. International Journal of Computer Vision 18 (1996), No. 1, pp. 41–60.

    Article  Google Scholar 

  7. Smith, S. M.: Asset-2: Real-time motion segmentation and object tracking. In: Proc. of the Fifth Int. Conf. on Computer Vision (1995), pp. 237–244.

  8. Trapp, R.: Entwurf einer Filterbank auf der Basis neurophysiologischer Erkenntnisse zur orientierungs-und frequenzselektiven Dekomposition von Bilddaten. Internal report NAVIS 01/96, Universität-GH-Paderborn. 1996.

  9. Wiklund, J., Granlund, G.: Tracking of multiple moving objects. In: Cappellini, V. (ed.): Time-varying image processing and moving object recognition, pp. 241–251. Amsterdam: Elsevier Science. 1987.

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Rights and permissions

Reprints and permissions

About this article

Cite this article

Hoischen, R., Mertsching, B. & Springmann, S. Object tracking in image sequences based on parametric features. Elektrotech. Inftech. 116, 390–394 (1999). https://doi.org/10.1007/BF03159200

Download citation

  • Issue Date:

  • DOI: https://doi.org/10.1007/BF03159200

Keywords

Schlüsselwörter

Navigation