2 Oct
2008
2 Oct
'08
10:58 p.m.
#162: Implement windows, neighborhoods, and corresponding iterators for complex-
based images
-----------------------+----------------------------------------------------
Reporter: levill_r | Owner: levill_r
Type: task | Status: assigned
Priority: major | Milestone: Olena 1.0
Component: Milena | Version: 1.0
Resolution: | Keywords:
-----------------------+----------------------------------------------------
Old description:
For a given (fixed) dimension ''n''
and a psite ''p'' on a ''n''-face,
implement windows/neighborhoods (and corresponding iterators) returning
the following elements.
~~`complex_lower_window`~~:: ~~the set of (''n''-1)-faces adjacent to
''p'' (using ``mln::p_faces`` and ``mln::faces_psite``?)~~ (mostly done
in r2459);
~~`complex_higher_window`~~:: ~~the set of (''n''+1)-faces adjacent to
''p'' (using ``mln::p_faces`` and ``mln::faces_psite``?)~~ (mostly done
in r2459);
`complex_lower_dim_connected_window`:: the set of ''n''-faces sharing
a
(''n''-1)-face with ''p'';
`complex_higher_dim_connected_window`:: the set of ''n''-faces sharing
a
(''n''+1)-face with ''p'';
`complex_cell_window`:: the set of the faces in the « cell » including
''p'' (named « ''p''-hat » in
couprie.08.pami), i.e. the set of all ''m''-faces adjacent to
''p'',
where ''m'' is in [0, ''n''-1].[[BR]]
In that definition, ''p'' is said adjacent to an
''m''-face ''q'' if
there is a sequence (''m,,1,,'', ''m,,2,,'', ...,
''m,,k,,'') of faces so
that
* ''m,,1,,'' is an (''n''-1)-face adjacent to
''p'' ;
* ''m,,2,,'' is an (''n''-2)-face adjacent to
''m,,1,,'' ;
* ... (and so on)
* ''m,,k,,'' is an (''m''+1)-face adjacent to
''q''.
`complex_boundary_cell_window`:: Likewise, but excluding ''p'' (named
«
''p''-hat* » in couprie.08.pami).
And maybe:
`complex_lower_dims_connected_window`:: the set of ''n''-faces sharing
a
(''n''-1)-face or (''n''-2)-face etc. (by transitivity)
with ''p'' (is it
useful?);
`complex_higher_dims_connected_window`:: the set of ''n''-faces
sharing
a (''n''+1)-face or (''n''+2)-face etc. (by transitivity)
with ''p'' (is
it useful?);
* what else?
As in #139,
* we might want to factor things using implementation classes:
* ``mln::internal::complex_vicinity``
* ``mln::internal::complex_vicinity_piter``
(we might even be able to factor them with graph-based ones);
* we could have one or several generic classes, using static or dynamic
predicates, to implement those windows and neighborhoods.
New description:
For a given (fixed) dimension ''n'' and a psite ''p'' on
a ''n''-face,
implement windows/neighborhoods (and corresponding iterators) returning
the following elements.
~~`complex_lower_window`~~:: ~~the set of (''n''-1)-faces adjacent to
''p'' (using ``mln::p_faces`` and ``mln::faces_psite``?)~~ (mostly done
in
r2459 and r2460);
~~`complex_higher_window`~~:: ~~the set of (''n''+1)-faces adjacent to
''p'' (using ``mln::p_faces`` and ``mln::faces_psite``?)~~ (mostly done
in
r2459 and r2460);
`complex_lower_dim_connected_window`:: the set of ''n''-faces sharing
a
(''n''-1)-face with ''p'';
`complex_higher_dim_connected_window`:: the set of ''n''-faces sharing
a
(''n''+1)-face with ''p'';
`complex_cell_window`:: the set of the faces in the « cell » including
''p'' (named « ''p''-hat » in
couprie.08.pami), i.e. the set of all ''m''-faces adjacent to
''p'',
where ''m'' is in [0, ''n''-1].[[BR]]
In that definition, ''p'' is said adjacent to an
''m''-face ''q'' if
there is a sequence (''m,,1,,'', ''m,,2,,'', ...,
''m,,k,,'') of faces so
that
* ''m,,1,,'' is an (''n''-1)-face adjacent to
''p'' ;
* ''m,,2,,'' is an (''n''-2)-face adjacent to
''m,,1,,'' ;
* ... (and so on)
* ''m,,k,,'' is an (''m''+1)-face adjacent to
''q''.
`complex_boundary_cell_window`:: Likewise, but excluding ''p'' (named
«
''p''-hat* » in couprie.08.pami).
And maybe:
`complex_lower_dims_connected_window`:: the set of ''n''-faces sharing
a
(''n''-1)-face or (''n''-2)-face etc. (by transitivity)
with ''p'' (is it
useful?);
`complex_higher_dims_connected_window`:: the set of ''n''-faces sharing
a
(''n''+1)-face or (''n''+2)-face etc. (by transitivity)
with ''p'' (is it
useful?);
* what else?
As in #139,
* we might want to factor things using implementation classes:
* ``mln::internal::complex_vicinity``
* ``mln::internal::complex_vicinity_piter``
(we might even be able to factor them with graph-based ones);
* we could have one or several generic classes, using static or dynamic
predicates, to implement those windows and neighborhoods.
Comment (by levill_r):
(Fix previous change.)
--
Ticket URL: <https://trac.lrde.org/olena/ticket/162#comment:6>
Olena <http://olena.lrde.epita.fr>
Olena, a generic and efficient C++ image processing library.