STYLE: Capitalize "Section" references in LaTeX for consistency

User Regex: 1:Capitalize before references.

Author: hjmjohnson

SoftwareGuide/Latex/DesignAndFunctionality/Registration.tex

@@ -78,7 +78,7 @@ \section{Registration Framework}
 function. Since the moving transform is shared between metric and optimizer,
 the above process will be repeated till the convergence criteria are met.
 
-Later in section~\ref{sec:FeaturesOfTheRegistrationFramework} you will get a
+Later in Section~\ref{sec:FeaturesOfTheRegistrationFramework} you will get a
 better understanding of the behind-the-scenes processes of ITKv4 registration
 framework. First, we begin with some simple registration examples.
 
@@ -182,7 +182,7 @@ \section{Features of the Registration Framework}%
 Only those cases may require invoking the \code{GetInverse()}
 method that most transforms offer. Before attempting this,
 read the examples on resampling illustrated in
-section~\ref{sec:GeometricalTransformationFilters} in order to
+Section~\ref{sec:GeometricalTransformationFilters} in order to
 familiarize yourself with the correct interpretation of the transforms.
 
 Now we come back to the situation illustrated in
@@ -232,7 +232,7 @@ \section{Multi-Modality Registration}%
 different groups~\cite{Collignon1995,Maes97,Viola1997}; a more detailed review
 can be found in~\cite{Hajnal2001,Pluim2003}. The Insight Toolkit currently
 provides two different implementations of Mutual Information metrics (see
-section~\ref{sec:Metrics} for details). The following example illustrates the
+Section~\ref{sec:Metrics} for details). The following example illustrates the
 practical use of one of these metrics.
 
 \subsection{Mattes Mutual Information}%
@@ -320,7 +320,7 @@ \subsection{Fundamentals}
 
 \section{Multi-Stage Registration}%
 \label{sec:MultiStageRegistration}
-In section~\ref{sec:MultiResolutionRegistration} you noticed how to tweak
+In Section~\ref{sec:MultiResolutionRegistration} you noticed how to tweak
 component settings between multi-resolution levels and saw how it can benefit the
 registration process. That is, the matching metric gets close to the optimal
 value before final parameter adjustments in full resolution.