Collaboration diagram for Markov Random Field-based Filters:
|
Classes | |
| class | itk::MRFImageFilter< TInputImage, TClassifiedImage > |
| Implementation of a labeller object that uses Markov Random Fields to classify pixels in an image data set. More... | |
| class | itk::RGBGibbsPriorFilter< TInputImage, TClassifiedImage > |
| RGBGibbsPriorFilter applies Gibbs Prior model for the segmentation of MRF images. The core of the method is based on the minimization of a Gibbsian energy function. This energy function f can be divided into three part: f = f_1 + f_2 + f_3; f_1 is related to the object homogeneity, f_2 is related to the boundary smoothness, f_3 is related to the constraint of the observation (or the noise model). The two force components f_1 and f_3 are minimized by the GradientEnergy method while f_2 is minized by the GibbsTotalEnergy method. More... | |
Detailed Description
Markov Random Field (MRF)-based Filters assume that the segmented image is Markovian in nature, i.e., adjacent pixels are likely to be of the same class. These methods typically combine intensity-based Filters with MRF prior models also known as Gibbs prior models.
