View source for Projects/Slicer3/2007 Project Week EdgeDetection

{|
|[[Image:ProjectWeek-2007.png|thumb|300px|Return to [[2007_Programming/Project_Week_MIT|Project Week Main Page]] ]]
|[[Image:BeforePicture.jpg|thumb|400px|BEFORE!]]
|}
{|
|[[Image:genuFA.jpg|thumb|300px|Output Image after edge detection.]]
|}


__NOTOC__
===Key Investigators===
* MIT: Elliot Uvero
* BWH: Lauren O'Donnell
* Kitware: Luis Ibanez

<div style="margin: 20px;">

<div style="width: 27%; float: left; padding-right: 3%;">

<h1>Objective</h1>

Enable edge detection of images in Slicer3.

</div>

<div style="width: 27%; float: left; padding-right: 3%;">

<h1>Approach, Plan</h1>

Create a Slicer3 CLP module for edge detection of images, using ITK filters.

</div>

<div style="width: 40%; float: left;">

<h1>Progress</h1>

A CLP module, Zero Crossing Based Edge Detection, has been developed in Slicer3.  With the great help of Luis, the documentation and error checking of the ITK ZeroCrossingEdgeDetectionFilter were improved, and the concept checking was improved to make sure that the pixel type is float or double.

</div>

<br style="clear: both;" />

<h1>Conclusions and Findings</h1>
* Concept checking should be set to ON in CMake when ITK is configured.
** When off (default) there is no check that input/output image is the same type for the itkZeroCrossingEdgeDetectionFilter.
* Currently the itkZeroCrossingEdgeDetectionFilter appears to function correctly with float and double input. 
** Short input produces different edge output.



<br style="clear: both;" />



{|
|[[Image:LuisAndElliot.jpg|thumb|320px|Working hard on ITK and Slicer3]]
|}