Mathematical study of the relaxed optical flow problem in the space BV (Ω)

被引:25
作者
Aubert, G [1 ]
Kornprobst, P [1 ]
机构
[1] UNSA, Lab JA Dieudonne, CNRS, UMR 6621, F-06108 Nice 2, France
关键词
measure theory; space of bounded variations; convex functions of measures; Gamma-convergence; elliptic equations; relaxation of ill-posed problems; optical flow; computer vision;
D O I
10.1137/S003614109834123X
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
This paper describes a variational approach for estimating a discontinuous optical flow from a sequence of images. Defined as the apparent motion of the image brightness pattern, the optical flow is very important in the computer vision community, where its accurate estimation is strongly needed. After a short overview of existing methods, we present a new variational method that we study in the space of bounded variations. We first present an integral representation of the optical flow problem which appears to be not lower semicontinuous. The relaxed functional is then calculated. We conclude by challenging questions about the possible numerical analysis of the abstract results.
引用
收藏
页码:1282 / 1308
页数:27
相关论文
共 53 条
[1]   ANALYSIS OF BOUNDED VARIATION PENALTY METHODS FOR ILL-POSED PROBLEMS [J].
ACAR, R ;
VOGEL, CR .
INVERSE PROBLEMS, 1994, 10 (06) :1217-1229
[2]  
AMBROSIO L, 1989, B UNIONE MAT ITAL, V3B, P857
[3]  
[Anonymous], WEAKLY DIFFERENTIABL
[4]  
[Anonymous], 1996, THESIS U NICE SOPHIA
[5]   THE EULER EQUATION FOR FUNCTIONALS WITH LINEAR GROWTH [J].
ANZELLOTTI, G .
TRANSACTIONS OF THE AMERICAN MATHEMATICAL SOCIETY, 1985, 290 (02) :483-501
[6]  
Anzellotti G., 1983, ANN MAT PUR APPL, V4, P293
[7]   A variational method in image recovery [J].
Aubert, G ;
Vese, L .
SIAM JOURNAL ON NUMERICAL ANALYSIS, 1997, 34 (05) :1948-1979
[8]  
AUBERT G, IN PRESS SIAM J APPL
[9]  
AUBERT G, 1997, IEEE T IMAGE PROCESS, V5, P298
[10]  
Aze D., 1986, Ric. Mat., V35, P125