Uncertainty characterization of the orthogonal Procrustes problem with arbitrary covariance matrices
详细信息 查看全文
- 作者：Pedro Lourenç ; oa ; plourenco@isr.tecnico.ulisboa.pt" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Bruno J. Guerreiroa ; b ; bguerreiro@isr.tecnico.ulisboa.pt" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Pedro Batistaa ; b ; pbatista@isr.tecnico.ulisboa.pt" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Paulo Oliveirad ; c ; a ; pjcro@isr.tecnico.ulisboa.pt" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Carlos Silvestree ; a ; csilvestre@umac.mo" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae
- 关键词：Weighted Procrustes statistics ; Perturbation theory ; Uncertainty characterization ; Map transformation
- 刊名：Pattern Recognition
- 出版年：2017
- 出版时间：January 2017
- 年：2017
- 卷：61
- 期：Complete
- 页码：210-220
- 全文大小：1667 K
文摘
- •
The uncertainty present in the n-dimensional Procrustes problem is characterized.
- •
Perturbation theory allows the determination of the covariance of the solution.
- •
The method is demonstrated for several dimensionalities on synthetic data.
- •
The effects of the number of points, baseline, and covariance shapes are detailed.