DocumentCode :
3508756
Title :
Image flow field detection
Author :
Fu, Shan ; Pridmore, Tony
Author_Institution :
Sch. of Eng., Sheffield Hallam Univ., UK
Volume :
2
fYear :
1996
fDate :
14-18 Oct 1996
Firstpage :
1090
Abstract :
In this paper, we suggest that the flow field should be smooth over the image, but in terms of its direction rather than its magnitude. A new smoothness constraint, on the image change direction, is proposed. An algorithm employing a multi-scale modification strategy with the proposed smoothness constraint used as a regularization term in an optimisation procedure is presented and used to estimate the image flow field. Experimental results show the algorithm to be efficient in dealing with both image expansion and discontinuous visible surfaces
Keywords :
estimation theory; image sequences; motion estimation; optimisation; discontinuous visible surfaces; image change direction; image expansion; image flow field detection; multi-scale modification strategy; optimisation procedure; regularization term; Cameras; Constraint optimization; Differential equations; Filters; Image segmentation; Image sequences; Motion estimation; Rendering (computer graphics);
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Signal Processing, 1996., 3rd International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-2912-0
Type :
conf
DOI :
10.1109/ICSIGP.1996.566283
Filename :
566283
Link To Document :
بازگشت