MetaMorphs: Deformable Shape and Texture Models

We present a new class of deformable models, MetaMorphs, whose formulation integrates both shape and interior texture. The model deformations are derived from both boundary and region information based on a variational framework. This framework represents a generalization of previous parametric and implicit geometric deformable models, by incorporating model interior texture information. The shape of the new model is represented implicitly as an "image" in the higher dimensional space of distance transforms. The interior texture is captured using a nonparametric kernel-based approximation of the intensity probability density function (p.d.f.) inside the model. The deformations that the model can undergo are defined using a space warping technique - the cubic B-spline based Free Form Deformations (FFD). When using the models for boundary finding in images, we derive the model dynamics from an energy functional consisting of both edge energy terms and texture energy terms. Thi...

• Type: Conference
• Year: 2004
• Where: CVPR
• Authors: Xiaolei Huang, Dimitris N. Metaxas, Ting Chen
• Keywords: Boundary Finding Robust, Computer Vision, CVPR 2004, Large Boundary Gaps, Model Deformations, Model Interior Texture, Texture Energy Terms