Abstract
The Monge–Kantorovich distance gives a metric between probability distributions on a metric space \({\mathbb X}\) and the MK distance is tied to the underlying metric on \({\mathbb X}\). The MK distance (or a closely related metric) has been used in many areas of image comparison and retrieval and thus it is of significant interest to compute it efficiently. In the context of finite discrete measures on a graph, we give a linear time algorithm for trees and reduce the case of a general graph to that of a tree. Next, we give a linear time algorithm which computes an approximation to the MK distance between two finite discrete distributions on any graph. Finally, we extend our results to continuous distributions on graphs and give some very general theoretical results in this context.
Similar content being viewed by others
References
Barahona F, Cabrelli CA, Molter UM (1992) Computing the Hutchinson distance by network flow methods. Random Comput Dyn 1:117–129
Barnsley M (1988) Fractals everywhere. Academic Press, New York
Boullion T, Odell P (1971) Generalized inverse matrices Wiley, New York
Brandt J, Cabrelli C, Molter U (1991) An algorithm for the computation of the Hutchinson distance. Inf Process Lett 40:113–117
Cabrelli CA, Molter UM (1995) The Kantorovich metric for probability measures on the circle. J Comput Appl Math 57:345–361
Chen K, Demko S, Xie R (1998) Similarity-based retrieval of images using color histograms. In: Yeung MM, Yeo BL, Bouman CA (eds) Storage and retrieval for image and video databases VII. Proceedings of SPIE, vol 3656, pp 643–652
Dantzig G, Thapa M (1997) Linear programming 1: introduction. Springer, New York
Deliu A, Mendivil F, Shonkwiler R (1991) Genetic algorithms for the 1-D fractal inverse problem. In: Proceedings of the 4th international conference on genetic algorithms, San Diego
Jang M, Kim S, Faloutsos C, Park S (2011) A linear-time approximation of the earth mover’s distance. In: Proceedings of the 20th ACM international conference on information and knowledge management, CIKM ’11, pp 504–514
Ling H, Okada K (2006) An efficient earth mover’s distance algorithm for robust histogram comparison. IEEE Trans PAMI 29(5):840–853
Nadler S (1992) Continuum theory: an introduction. Dekker, New York
Rachev ST, Ruschendorf L (1998) Mass transportation problems, vol I. Theory. Springer, New York
Rubner Y, Tomasi C, Guibas J (2000) The earth mover’s distance as a metric for image retrieval. Int J Comput Vis 40(2):99–121
Shirdhonkar S, Jacobs D (2008) Approximate earth mover’s distance in linear time. In: Proceedings of the IEEE international conference on computer vision and pattern recognition, CVPR
Acknowledgments
The author would like to thank Steve Demko for a very large number of illuminating discussions on this topic and explaining to me his work in Chen et al. (1998). This work is partially supported by a Discovery Grant from the Natural Sciences and Engineering Research Council of Canada (238549).
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by Jinyun Yuan.
Rights and permissions
About this article
Cite this article
Mendivil, F. Computing the Monge–Kantorovich distance. Comp. Appl. Math. 36, 1389–1402 (2017). https://doi.org/10.1007/s40314-015-0303-7
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s40314-015-0303-7
Keywords
- Monge-Kantorovich distance
- Quotient norm
- Graph quotient
- Matrix pseudo-inverse
- Digital image comparison
- Earth-mover’s distance