Statistical Image Sequence Processing for Temporal Change Detection

  • Martin Brocke
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2449)


The aim is to detect sudden temporal changes in image sequences, focusing on bright objects that appear in a few consecutive frames. The proposed algorithm detects such outliers by computing a variance weighted deviation from mean values for every pixel. On this result, an object segmentation based on 2D-moments and its invariants is done frame by frame at a ≈ 3σ threshold. The algorithm was designed for a wide range of tasks in pre-processing as a tool for detection of fast temporal changes such as suddenly appearing or moving objects. Two different applications on noisy sequence data were realized. The entire system proved to fulfill the requirements of industrial environments for online process control and scientific demands for data rejection.


change detection outliers image sequences process control automation 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Aach, T., Kaup, A., Mester, R.: Bayesian Algorithms for Change Detection in Image Sequ. Using Markov Random Fields, Sig. Proc.-IC, vol. 7, no. 2, pp. 147–160, 1995Google Scholar
  2. 2.
    Beckman, R.J., Cook, R.D.: Outlier....s, Technometrics, vol. 25, pp. 119–149, 1983zbMATHCrossRefMathSciNetGoogle Scholar
  3. 3.
    Bouthemy, P., Lalande, P.: Detection and tracking of moving objects based on a statistical regularization method in space and time, Proc. ECCV, pp. 307–311, 1990Google Scholar
  4. 4.
    Brocke M., Schmidt, M. D. et al.: Verfahren zur autom. Beurteilung von Laserverarbeitungsprozessen, German Patent, DE 10103255, disclosure: 25.7.2002Google Scholar
  5. 5.
    Garbe, C., Jähne, B.: Reliable estimates of the sea surface heatflux from image sequences, In: B. Radig, Mustererkennung 2001, 23. DAGM-Symp., MünchenGoogle Scholar
  6. 6.
    Hötter, M., Mester, R., Meyer, M.: Detection of moving objects in natural scenes, Proc. Carnahan Conf. on Security Technology, pp. 47–52, 1995Google Scholar
  7. 7.
    Hsu, Y. Z., Nagel, H. H. and Rekers, G.: New likelihood test methods for change detection in image sequences, CVGIP, vol. 26, pp. 73–106, 1984Google Scholar
  8. 8.
    Kreßel U. et al.: Polynomklassifikator vs. Multilayer-Perzeptron, In: R.E. Großkopf: Mustererkennung 1990, 12. DAGM-Symp. OberkochenGoogle Scholar
  9. 9.
    Madalla, G. S., Yin Y.: Outliers, unit roots and robust estimation, In: G. S. Madalla et al: Handbook of Statistics, vol. 15, Elsevier, 1997Google Scholar
  10. 10.
    Nordbruch, S. et al.: Analyse von HDRC-Bildern des Werkstoffübergangs des MSG-Schweißprozesses, In: G. Sommer, Mustererkennung 2000, 22. DAGM-Symp., KielGoogle Scholar
  11. 11.
    Sethi, I., Patel, N.: A Statistical Approach to Scene Change Detection, SPIE Conf. Stor. and Retr. for Image and Video Database III, no. 2420, pp. 329–338Google Scholar
  12. 12.
    Teague, M. R.: Image Analysis via the General Theory of Moments, Opt. Soc. of America, vol. 70, no. 8, pp. 920–930, 1980MathSciNetCrossRefGoogle Scholar
  13. 13.
    Thompson, W. R.: On a criterion for the rejection of observations, Ann. of Math. Stat., vol. 6, pp. 214–219, 1935zbMATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Martin Brocke
    • 1
    • 2
  1. 1.Robert Bosch GmbH, FV/PLF2Stuttgart
  2. 2.Interdisciplinary Center for Scientific Computing, INF 368Heidelberg

Personalised recommendations