Sc. Pei et Lg. Liou, USING MOMENTS TO ACQUIRE THE MOTION PARAMETERS OF A DEFORMABLE OBJECTWITHOUT CORRESPONDENCES, Image and vision computing, 12(8), 1994, pp. 475-485
Citations number
19
Categorie Soggetti
Computer Sciences, Special Topics",Optics,"Engineering, Eletrical & Electronic","Computer Science Artificial Intelligence","Computer Science Software Graphycs Programming","Computer Science Theory & Methods
In the topic of motion analysis in image understanding, we usually nee
d to find the motion of a particular object. However, frequently the c
orrespondence problem, which is well known to be difficult, must first
be solved. If the correspondence problem can be avoided in motion ana
lysis, it will greatly reduce the difficulty of solving motion paramet
ers. Using the concept of object normalization described in this paper
, we can determine the motion parameters without solving the correspon
dence problem. The object of interest is not necessarily assumed rigid
here. It can undergo a uniformly, linearly deformable motion (affine
transformation), including translation, rotation and skewing. Under th
e assumption of this motion model, a closed-form solution of the desir
ed motion parameters can be obtained efficiently by our proposed metho
d. In some special cases where the closed-form solution cannot be dete
rmined uniquely, a modified method is also developed to solve them. A
theoretical error analysis is presented to show for what kinds of obje
ct it is inherently hard to acquire motion parameters.