** Next:** Comparing Registration Algorithms
** Up:** Experimental Validation
** Previous:** Validation using Warped Images

##

Sensitivity

The size of perturbation that can be detected in the validation
experiments will depend both on the change in the values of the
measures as a function of misregistration and the standard error
of those values. To quantify this, we define the sensitivity of a
measure as follows.

where 84#84 is the value of the measure for some degree
of deformation 79#79,
85#85 is the standare error of the
estimate of 84#84. 86#86 is the change in 79#79 required for
84#84 to change by one noise standard error, which indicates the
lower limit of change in misregistration 79#79 which can be detected
by the measure. 87#87 is a function of 79#79; to simplify comparison
between different methods of evaluation, we also use the mean
sensitivity over a range of values of 79#79.

In order to compare the sensitivities of different methods of
evaluation, we need also to estimate the expected error in 87#87.
Since the validation experiments provided repeated estimates of
84#84, we can obtain empirical estimates of the errors in 84#84,
88#88, and
85#85. These can be combined, using error
propagation, to estimate the uncertainty in the estimate of
sensitivity.

** Next:** Comparing Registration Algorithms
** Up:** Experimental Validation
** Previous:** Validation using Warped Images
Roy Schestowitz
2007-03-11