Next: Construction of a RJMCMC
Up: Roads Extraction using a
Previous: Stochastic model for a
Data term for the Candy model
Hypothesis tests :
H_{3} : we have three different regions
Figure 5:
Three region mask

Using a Gaussian assumption, the log of the likelihood function is :
H_{2} : we have two different
regions
Figure 6:
Two region mask

The log of the likelihood function :
H_{1} : the segment is in the middle of an homogeneous region
Figure 7:
One region mask

The log of the likelihood function :

= 

(16) 
The Total Energy for the Candy model
The conditional energy for a segment :
U_{D}(s) 
= 





(17) 
Depending on the type of image, we may add to the conditional term :

(18) 
or :

(19) 
The total energy becomes :
U(S/D)=U_{D}(S)+U(S)

(20) 
with :

(21) 
Next: Construction of a RJMCMC
Up: Roads Extraction using a
Previous: Stochastic model for a
Radu Stoica
20000417