Hidden MRF detection of motion of objects with uniform brightness

  • Adam Kuriański
  • Mariusz Nieniewski
Part of the Lecture Notes in Computer Science book series (LNCS, volume 974)


The hidden Markov Random Field (MRF) model for motion detection in image sequences is described. A typical MRF model uses two observations: the difference in brightness between two consecutive images, and the value obtained from the mask of temporal changes. The performance of this model can be improved by including the third observation: the brightness at a given pixel. The paper gives the necessary equations and presents an example of motion detection of the object with uniform brightness.



brightness value at pixel (i,j) of the k-th image


brightness difference o k (i,j)=fk+1(i,j)−f k (i,j)


value at pixel (i, j) of the mask of temporal changes between k-th and (k+1)-th image

C={1, 0}

set of values which can be assigned to a particular pixel (i,j) by the mask of temporal changes


label assigned to pixel (i,j) of the k-th image by the mask of moving objects

L={a, b}

set of labels which can be assigned to a particular pixel by the mask of the moving object, Superscript T denotes the transposition of a vector.


  1. 1.
    Besag, J.: On the statistical analysis of dirty pictures. Journ. Royal Statist. Soc. 48 series B (1986) 259–302Google Scholar
  2. 2.
    Bouthemy, P., Lalande, P.: Detection and tracking of moving objects based on a statistical regularization method in space and time. Proc. European Conf. Comp. Vision, Antibes (1990) 307–311Google Scholar
  3. 3.
    Geman, S., Geman, D.: Stochastic relaxation, Gibbs distributions and the Bayesian restoration of images. Pattern Anal. Mach. Int. 6 (1984) 721–741Google Scholar
  4. 4.
    Lalande, P., Bouthemy, P.: A statistical approach to the detection and tracking of moving objects in an image sequence. Proc. 5th ESPC EUSIPCO 90, Barcelona (1990)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1995

Authors and Affiliations

  • Adam Kuriański
    • 1
  • Mariusz Nieniewski
    • 1
    • 2
  1. 1.Institute of Fundamental Technological ResearchPASWarsaw
  2. 2.Dept. of Fundamental Research in Electrical EngineeringPASWarsaw

Personalised recommendations