Multispectral Random Field Models for Synthesis and Analysis of Color Images
IEEE Transactions on Pattern Analysis and Machine Intelligence
Maximum Likelihood Estimation Methods for Multispectral Random Field Image Models
IEEE Transactions on Pattern Analysis and Machine Intelligence
| Hi-index | 0.00 |
A novel model-based approach for estimation of the velocity of crack growth from microfractographical images is proposed. These images are represented by a Gaussian Markov random field model and the crack growth rate is modelled by a linear regression model in the Gaussian-Markov parameter space. The method is numerically very efficient because both crack growth rate model parameters as well as the underlying random field model parameters are estimated using fast analytical estimators.