A new approach for regularization of inverse problems in images processing

Innocent Souopgui 1 Emmanuel Kamgnia 2 François-Xavier Le Dimet 1 Arthur Vidard 1
1 MOISE - Modelling, Observations, Identification for Environmental Sciences
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : Optical flow motion estimation from two images is limited by the aperture problem. A method to deal with this problem is to use regularization techniques. Usually, one adds a regularization term with appriopriate weighting parameter to the optical flow cost funtion. Here, we suggest a new approach to regularization for optical flow motion estimation. In this approach, all the regularization informations are used in the definition of an appropriate norm for the cost function via a trust function to be defined, one does not ever need weighting parameter. A simple derivation of such a trust function from images is proposed and a comparison with usual approaches is presented. These results show the superiority of such approach over usual ones.
Document type :
Conference papers
Complete list of metadatas

Contributor : Innocent Souopgui <>
Submitted on : Wednesday, November 24, 2010 - 6:07:53 PM
Last modification on : Wednesday, April 11, 2018 - 1:58:31 AM
Long-term archiving on: Friday, February 25, 2011 - 3:00:27 AM



  • HAL Id : hal-00539635, version 1



Innocent Souopgui, Emmanuel Kamgnia, François-Xavier Le Dimet, Arthur Vidard. A new approach for regularization of inverse problems in images processing. CARI'10 - 10th African Conference on Research in Computer Science and Applied Mathematics, Oct 2010, Yamoussoukro, Côte d’Ivoire. pp.173-180. ⟨hal-00539635⟩



Record views


Files downloads