DocumentCode
1121886
Title
Motion and structure from point correspondences with error estimation: planar surfaces
Author
Weng, Juyang ; Ahuja, Narendra ; Huang, Thomas S.
Author_Institution
Illinois Univ., Urbana, IL, USA
Volume
39
Issue
12
fYear
1991
fDate
12/1/1991 12:00:00 AM
Firstpage
2691
Lastpage
2717
Abstract
The determination of motion and structure of a planar scene from monocular image sequences, is studied. A new and simpler linear algorithm, that gives closed-form solutions for motion and structure parameters using point correspondences between two images, assuming that the coplanar points undergo a rigid motion in 3-D space, is presented. A series of analytical results is established. From two perspective views of a planar scene which is undergoing a rigid motion, there are generally two (normalized) interpretations for motion parameters and the positions of the object plane. These two interpretations, one vertical and the other illusive, are both valid in the sense that they render the same pair of images. The authors identify all the special cases in which the number of interpretations is not two, and derive necessary and sufficient geometrical conditions for those special cases to occur. The approach to error estimation is based on the first-order perturbation. The estimated errors provide quantitative assessment for the accuracy of the solutions. They also indicate degenerate or nearly degenerate configurations in the presence of noise
Keywords
error analysis; estimation theory; picture processing; 3-D space; closed-form solutions; coplanar points; error estimation; first-order perturbation; geometrical conditions; linear algorithm; monocular image sequences; motion parameters; noise; object plane; planar scene; planar surfaces; point correspondences; structure parameters; Closed-form solution; Ear; Error analysis; Image edge detection; Image motion analysis; Image sequences; Layout; Least squares approximation; Motion analysis; Motion estimation;
fLanguage
English
Journal_Title
Signal Processing, IEEE Transactions on
Publisher
ieee
ISSN
1053-587X
Type
jour
DOI
10.1109/78.107418
Filename
107418
Link To Document