Title :
Estimation of 2D motion field based on affine geometric invariance
Author_Institution :
Signal Process. Lab., DSO Nat. Labs., Singapore
Abstract :
A novel algorithm of measuring visual motion induced by egomotion of camera was developed from two views. In this algorithm, by using the geometric invariance, the correspondence of any image point can be obtained by solving a system of 5 linear equations. Using the second order cone programming the ill-conditioned least squares produced by utilisation of bilinear estimation scheme was successfully solved. The validity of the approach is verified by the experiment results
Keywords :
computational geometry; computer vision; invariance; mathematical programming; motion estimation; 2D motion estimation; affine geometric invariance; computer vision; cone programming; image point; least squares; Cameras; Equations; Laboratories; Layout; Least squares methods; Motion estimation; Motion measurement; Robustness; Signal processing algorithms; Working environment noise;
Conference_Titel :
Pattern Recognition, 2000. Proceedings. 15th International Conference on
Conference_Location :
Barcelona
Print_ISBN :
0-7695-0750-6
DOI :
10.1109/ICPR.2000.903723
