@@ -246,44 +246,20 @@ \subsection{Mattes Mutual Information}
246246% \input{ImageRegistrationHistogramPlotter.tex}
247247
248248
249- \section { Centered Transforms }
249+ \section { Center Initialization }
250250
251251The ITK image coordinate origin is typically located in one of the image
252252corners (see the Defining Origin and Spacing section of Book 1 for details).
253253This results in counter-intuitive transform behavior when rotations and scaling
254254are involved. Users tend to assume that rotations and scaling are performed
255- around a fixed point at the center of the image. In order to compensate for
256- this difference in natural interpretation, the concept of \emph {centered }
257- transforms have been introduced into the toolkit. The following sections
258- describe the main characteristics of such transforms.
259-
260- The introduction of the centered transforms in the Insight Toolkit reflects the
261- dynamic nature of a software library when it evolves in harmony with the
262- requests of the community that it serves. This dynamism has, as everything else
263- in real life, some advantages and some disadvantages. The main advantage is that
264- when a need is identified by the users, it gets implemented in a matter of days
265- or weeks. This capability for rapidly responding to the needs of a community
266- is one of the major strengths of Open Source software. It has the additional
267- safety that if the rest of the community does not wish to adopt a particular
268- change, an isolated user can always implement that change in her local copy of
269- the toolkit, since all the source code of ITK is available in a Apache 2.0
270- license\footnote {\url {http://www.opensource.org/licenses/Apache-2.0}} that
271- does not restrict modification nor distribution of the code, and that does not
272- impose the assimilation demands of viral licenses such as
273- GPL\footnote {\url {http://www.gnu.org/copyleft/gpl.html}}.
274-
275- The main disadvantage of dynamism, is of course, the fact that there is
276- continuous change and a need for perpetual adaptation. The evolution of
277- software occurs at different scales, some changes happen to evolve in localized
278- regions of the code, while from time to time accommodations of a larger scale
279- are needed. The need for continuous changes is addressed in Extreme Programming
280- with the methodology of \emph {Refactoring }. At any given point, the structure
281- of the code may not project the organized and neatly distributed architecture
282- that may have resulted from a monolithic and static design. There are, after
283- all, good reasons why living beings can not have straight angles. What you are
284- about to witness in this section is a clear example of the diversity of species
285- that flourishes when evolution is in action~\cite {Darwin1999 }.
286-
255+ around a fixed point at the center of the image. In order to compensate for
256+ this difference in expected interpretation, the concept of \emph {center } of
257+ transform has been introduced into the toolkit. This parameter is
258+ generally a \emph {fixed } parameter that is not optimized during
259+ registration, so initialization is crucial to get efficient and
260+ accurate results. The following sections describe the main
261+ characteristics and effects of initializing the center of a
262+ transform.
287263
288264\subsection {Rigid Registration in 2D }
289265\label {sec:RigidRegistrationIn2D }
@@ -308,7 +284,7 @@ \subsection{Rigid Transform in 3D}
308284
309285
310286
311- \subsection {Centered Affine Transform }
287+ \subsection {Centered Initialized Affine Transform }
312288\label {sec:CenteredAffineTransform }
313289\input {ImageRegistration9.tex }
314290
0 commit comments