Deformable models direct supervised guidance: A novel paradigm for automatic image segmentation
详细信息 查看全文
- 作者：Nicola Bovaa ; c ; nbova@inf.ed.ac.uk" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Viktor Gá ; lb ; viktor.gal@ugent.be" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Ó ; scar Ibá ; ñ ; eza ; d ; oscar.ibanez@decsai.ugr.es" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Ó ; scar Cordó ; na ; d ; ocordon@decsai.ugr.es" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae
- 关键词：Image segmentation ; Deformable models ; Level set ; Machine learning ; Optimization ; Direct guidance
- 刊名：Neurocomputing
- 出版年：2016
- 出版时间：12 February 2016
- 年：2016
- 卷：177
- 期：Complete
- 页码：317-333
- 全文大小：11499 K
文摘
- •
Defining a priori the energy function guiding deformable models evolution is hard.
- •
We show an alternative, fully automatic framework for deformable model adjustment.
- •
The deformable model adjustment is directly guided by a machine learning method.
- •
Learning is driven by the ground truth information provided in the training set.
- •
Our framework is composed by four customizable parts plus an integration mechanism.