📄 help_imare.html
字号:
<td BGCOLOR="#3366FF"> <font color="#3366FF">**</font></td>
</tr>
<tr BGCOLOR="#FFCC99">
<td BGCOLOR="#3366FF"> <font color="#3366FF">**</font></td>
<td BGCOLOR="#99FF99"> <font color="#99FF99">**</font></td>
<td> <font color="#FFCC99">**</font></td>
<td><font color="#FFCC99">**</font></td>
<td><font color="#FFCC99">**</font></td>
<td BGCOLOR="#99FF99"> <font color="#99FF99">**</font></td>
<td BGCOLOR="#3366FF"> <font color="#3366FF">**</font></td>
</tr>
<tr BGCOLOR="#99FF99">
<td BGCOLOR="#3366FF"> <font color="#3366FF">**</font></td>
<td> <font color="#99FF99">**</font></td>
<td><font color="#99FF99">**</font></td>
<td><font color="#99FF99">**</font></td>
<td> <font color="#99FF99">**</font></td>
<td><font color="#99FF99">**</font></td>
<td BGCOLOR="#3366FF"> <font color="#3366FF">**</font></td>
</tr>
<tr BGCOLOR="#3366FF">
<td><font color="#3366FF">**</font></td>
<td><font color="#3366FF">**</font></td>
<td><font color="#3366FF">**</font></td>
<td> <font color="#3366FF">**</font></td>
<td><font color="#3366FF">**</font></td>
<td> <font color="#3366FF">**</font></td>
<td> <font color="#3366FF">**</font></td>
</tr>
</table></center>
<br>
<br>
<ul>
<li>
<font color="#000099">choice
which invariants from INV were used to form INVS</font></li>
</ul>
</ul>
</ul>
<p><br><b><font color="#000099">Output of computed CPs and invariants</font></b>
<br><b><font color="#993300">
<Saving CP, INV></font></b>
<br><font color="#000099">The CPs are saved in both PGM (0-1 matrix) and
MAT (x and y coordinates and the filename of corresponding image file,
where points were detected). The computed invariants together with used
parameters are saved in MAT format.</font>
<br>
<p><b><font color="#000099">Output of transform parameters and transformed
sensed image</font></b>
<br><b><font color="#993300">
<Saving transform></font></b>
<br><font color="#000099">The computed parameters of the registration together
with the transformed Sensed image are saved to the MAT file. Transformed
Sensed image is saved as well in TIFF format.</font>
<p>
<hr SIZE=3 WIDTH="100%">
<br><a NAME="viz"></a>
<br><b><font color="#72BF64"><font size=+1>Visualization</font></font></b>
<p><font color="#000099">IMARE toolbox offers some methods to visualize/check
achieved results. They allow to show detected CPs, computed INVs etc.</font>
<br>
<p><b><font color="#000099">Clearing images</font></b>
<br><b><font color="#993300">
<Clear></font></b>
<br><font color="#000099">The way how clean up all the points and values
plotted in image windows. Everything except the images is deleted</font>
<br>
<p><b><font color="#000099">Zooming</font></b>
<br><b><font color="#993300">
<Zoom></font></b>
<br><font color="#000099">The way how to zoom in details in the images.
It could be switched on or off. Left mouse zooms in, right mouse zooms
out.</font>
<br>
<p><b><font color="#000099">Coordinates of detected control points</font></b>
<br><b><font color="#993300">
<[X,Y]></font></b>
<br><font color="#000099">It shows window with two lists of CP coordinates,
one from Reference and one from Sensed images. The window can be closed
by clicking on OK button.</font>
<br>
<p><b><font color="#000099">Locations of detected control points</font></b>
<br><b><font color="#993300">
<CP +></font></b>
<br><font color="#000099">It shows detected CPs in Reference and Sensed
images as red spots.</font>
<br>
<p><b><font color="#000099">Showing invariant values</font></b>
<br><b><font color="#993300">
<INV values></font></b>
<br><font color="#000099">It allows to show computed invariants. You can
select CP which you want to check by mouse clicking nearby the point red
mark. The window with the number of CP, its coordinates and values of invariants
(maximum first 8 invariants in the case of <b>rotational invariants</b>)
is activated. At the beginning, Reference image area is active. To switch
for the Sensed image, press "<b>S</b>" key, to go back press "<b>R</b>"
key. To quit press "<b>Q</b>" key. If invariants were computed even for
points in close neighborhood of chosen CP, values for all points are shown
in list boxes. Here, you can display values of certain invariant for all
points or you can choose one point from the neighborhood and check corresponding
invariants.</font>
<p> <p>
<p><b><font color="#000099">Detected control points used for invariant
computation</font></b>
<br><b><font color="#993300">
<CP + INV></font></b>
<br><font color="#000099">It marks with green circles these CPs, for which
the invariants were computed. The CPs which are too close to the border
of the image are excluded.</font>
<br><font color="#000099">The critical distance is equal to the radius
of the neighborhood from which invariants are computed.</font>
<br>
<hr SIZE=3 WIDTH="100%">
<p><a NAME="actions"></a><b><font color="#72BF64"><font size=+1>Actions
and parameters setting</font></font></b>
<p><font color="#000099">The toolbox allows either to process whole image
registration process or you can just do CP detection and/or invariant computation.
Following section describes how to achieve these steps and explains the
impact of parameters which can be chosen.</font>
<p>
<hr SIZE=3 WIDTH="100%">
<br><a NAME="CPdet"></a><b><font color="#72BF64"><font size=+1>Control
point detection</font></font></b>
<br><font color="#000099">For control point detection, three methods are
available.</font>
<ul>
<li>
<font color="#000099"><b>Manual detection</b> - allows user to detect points
manually in both images</font></li>
</ul>
<ul><font color="#000099">Sequence of steps:</font>
<p><b><font color="#663300"> <CP detection method>
</font></b><font color="#000099">... <b>Manual</b></font>
<p><b><font color="#663300"> <Number of points>
</font></b><font color="#000099">... enter the desired
number of points to be detected</font>
<li>
<p align="left"><font color="#000099"> </font><b><font color="#663300"><CP
detection>
</font></b><font color="#000099">... activates interactive
process of point selection. It starts at the Reference image area, by left
mouse clicking</font>
<br><font color="#000099">
you can select point, it is marked by red cross. If this is the correct
choice, right mouse clicking confirms</font>
<br><font color="#000099">
the choice and the selected point is marked by blue cross and number. Otherwise
you can continue by left mouse</font> <font color="#000099">clicking till the
correct position is selected. At the message line the number of already
marked points is shown.</font>
<br>
<br>
<br>
</li>
<li>
<font color="#000099"><b>Corner similarity method</b> - allows user automatically
detect points by means of the method, described in</font></li>
</ul>
<i><font color="#000099">
Zitova B., Kautsky J., Peters G., Flusser J.: <a href="http://www.utia.cas.cz:80/cgi-bin/toASCII/library/prace/990009.pdf" TARGET="_blank">
Robust detection of significant points in multiframe images</a>. Pattern Recognition Letters, 20</font></i>
<br><i><font color="#000099">
(1999),2, 199-206.</font></i>
<ul><font color="#000099">Sequence of steps:</font>
<ul> </ul>
<b><font color="#663300"> <CP detection method>
</font></b><font color="#000099">... <b>Corner similarity</b></font>
<p><b><font color="#663300"> <Number of points>
</font></b><font color="#000099">... enter the desired
number of points to be detected</font>
<br><b><font color="#663300"> <Min distance>
</font></b><font color="#000099">... enter the minimum
allowed distance between points to be detected</font>
<br><b><font color="#663300"> <Mean radius>
</font></b><font color="#000099">... enter the radius
for the mean value plane computation (see the article), default value <b>2</b></font>
<br><b><font color="#663300"> <Sign change radius>
</font></b><font color="#000099">... enter the radius
for the sign change computation (see the article), default value <b>4</b></font>
<br><b><font color="#663300"> <Straight>
</font></b><font color="#000099">... enter the allowed
error in the straightness of lines (see the article), default value <b>1</b></font>
<br><b><font color="#663300"> <Angle>
</font></b><font color="#000099">... enter the allowed
divergence from 90<sup>o </sup>(see the article), default value <b>7</b></font>
<br><b><font color="#663300"> <Distance straight>
</font></b><font color="#000099">... enter the critical
distance from straight lines (see the article), default value <b>3</b></font>
<p><font color="#000099"> </font><b><font color="#663300"><CP
detection>
</font></b><font color="#000099">... activates automatic
process of control point selection.</font></ul>
<ul>
<br>
<li>
<font color="#000099"><b>Harris method</b> - allows user automatically
detect points by means of the Harris method</font></li>
<br>
<p>
<p><font color="#000099">Sequence of steps:</font></ul>
<ul><b><font color="#663300"> <CP detection method>
</font></b><font color="#000099">... <b>Harris</b></font>
<p><b><font color="#663300"> <Number of points>
</font></b><font color="#000099">... enter the desired
number of points to be detected</font>
<br><b><font color="#663300"> <Min distance>
</font></b><font color="#000099">... enter the minimum
allowed distance between points to be detected</font>
<br><b><font color="#663300"> <STD of Gaussian>
</font></b><font color="#000099">... enter the STD of
Gaussian used for the point computation</font>
<p><font color="#000099"> </font><b><font color="#663300"><CP
detection>
</font></b><font color="#000099">... activates automatic
process of control point selection.</font></ul>
<font color="#000099">After the detection process ends, located CPs are
marked by red spots in both images. Points can be saved (coordinates X
0-1 logical matrix) by means of</font>
<br><b><font color="#993300"><Saving CP, INV>. </font></b><font color="#000099">Points
can be visualized in future by means of </font><b><font color="#993300"><[X,Y]>,
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -