We describe an algorithm for locally-adaptive parameter estimation of spatially inhomogeneous Markov random elds (MRFs). In particular, we establish that there is a unique solution which maximizes the local pseudo-likelihood in the inhomogeneous MRF model. Subsequently we demonstrate how Besag's iterative conditional mode (ICM) procedure can be generalized from homogeneous MRFs to inhomogeneous...

diagnosis of corneal diseases is possible by measuring and evaluation of corneal thickness in different layers. thus, the need for precise segmentation of corneal layer boundaries is inevitable. obviously, manual segmentation is time‑consuming and imprecise. in this paper, the gaussian mixture model (gmm) is used for automatic segmentation of three clinically important corneal boundaries on opt...

It is known that parameter selection for data sampling frequency and segmentation techniques (including different methods and window sizes) has an impact on the classification accuracy. For Ambient Assisted Living (AAL), no clear information to select these parameters exists, hence a wide variety and inconsistency across today's literature is observed. This paper presents the empirical investig...

Automatic segmentation of stroke lesions in magnetic resonance imagery is a difficult problem because anatomical knowledge is required for the most accurate decisions. Without such knowledge, classification rules seem inconsistent. We propose a hybrid boundary and region based segmentation model built upon nonlinear scalespace and geometric active contours that captures the various segmentation...

A common approach to analysis of mouse behavior recorder by video tracking systems employs manual segmentation and labeling of mouse activity into behavioral acts. Developed automatic methods allow segmentation only to lingering and progression segments, suffer from poor precision and require parameter tuning. We propose a novel approach based on hidden Markov model for simultaneous segmentatio...

In the isointense stage, the accurate volumetric image segmentation is a challenging task due to the low contrast between tissues. In this paper, we propose a novel very deep network architecture based on densely convolutional network for volumetric brain segmentation. The proposed network architecture provides a dense connection between layers that aims to improve the information flow in the n...

We use Bayesian methods to design cellular neural networks for signal processing tasks and the Boltzmann Machine learning rule for parameter estimation. The learning rule can be used for models with uhidden" units, or for compietely unsupervised learning. The latter is exemplified by unsupervised adaptation of an image segmentation cellular network, in particular we apply the learning rule to a...

Statistical shape models are powerful tools for model-based segmentation and have been successfully applied to the segmentation of various structures in medical images. Though the segmentation algorithms based on statistical shape models are simple, finding corresponding landmarks for the construction of the models is a challenging optimisation task. State-of-the-art algorithms that solve the c...

This paper presents a novel method of unsupervised segmentation for synthetic aperture radar (SAR) images. Firstly, we define a generalized multiresolution likelihood ratio (GMLR), which classifies different kinds of signals more accurately than classical likelihood ratio by fusing more and different signal features. For our SAR image segmentation application, multiresolution stochastic structu...