Title :
Motion estimation based on Markov random fields
Author :
Rouchouze, B. ; Mathieu, P. ; Gaidon, T. ; Barlaud, M.
Author_Institution :
Nice Univ., Valbonne, France
Abstract :
This paper deals with a new method for estimation of motion field in image sequence coding domain. The proposed method is based on a pel-recursive technique and characterised by two fundamental points: the use of a convex function for the regularization and a deterministic relaxation algorithm for edge preserving regularization. These two points are completely new in motion estimation techniques
Keywords :
Markov processes; deterministic algorithms; edge detection; image coding; image sequences; maximum likelihood estimation; minimisation; motion estimation; recursive estimation; MAP; Markov random fields; convex function; deterministic relaxation algorithm; edge preserving regularization; image sequence coding domain; motion estimation; pel-recursive technique; regularization; HDTV; Image coding; Image motion analysis; Image sequences; Laboratories; Markov random fields; Motion estimation; Multimedia systems; Optical computing; Videoconference;
Conference_Titel :
Image Processing, 1994. Proceedings. ICIP-94., IEEE International Conference
Conference_Location :
Austin, TX
Print_ISBN :
0-8186-6952-7
DOI :
10.1109/ICIP.1994.413845
