Paper Info
Title
Randomized motion estimation
Abstract
Motion estimation is known to be a non-convex optimization problem. This non-convexity comes from several ambigui- ties in motion estimation such as the aperture problem, or fast motion relative to the magnitude of the image gradient. In this paper, we propose a fast random search algorithm to es- timate motion. Randomized algorithms are very popular in computer science and optimization for non-convex problems. However, to the best of our knowledge none has been used so far for motion estimation, due to complexity constraints. In this paper, we propose two fast algorithms to perform ran- dom search on image pixels. One produces a dense optical flow by matching patches. The other one takes advantage of a quad tree or segmentation tree structure of the image to es- timate motion in regions of increasing size. Quantitative and visual results show that the motion obtained seems to be a very advantageous compromise between speed and quality of estimated motion.
Year
DOI
Venue
2010
10.1109/ICIP.2010.5652514
Image Processing
Keywords
Field
DocType
concave programming,motion estimation,quadtrees,image gradient,image pixel,motion estimation,nonconvex optimization,quad tree structure,random search algorithm,segmentation tree structure,Matching,Optical flow,Random search,Segmentation tree
Structure from motion,Random search,Computer vision,Image gradient,Motion field,Quarter-pixel motion,Pattern recognition,Computer science,Image segmentation,Artificial intelligence,Motion estimation,Optical flow
Conference
ISSN
ISBN
Citations 
1522-4880 E-ISBN : 978-1-4244-7993-1
978-1-4244-7993-1
2
PageRank 
References 
Authors
0.43
11
2
Name
Order
Citations
PageRank
Sylvain Boltz1465.61
Frank Nielsen21256118.37