__header__
The edge finding tools finds edge pixels. The method employed is directional and sub pixel edges are found. The number of points found depends on the parameter settings. An edge is defined as a transition from light to dark or dark to light intensities.
The tool can fine gradient edges or threshold edges.
Drag the Edge Finding Tool onto a program to display its ROI on the viewer.
Position the ROI over the edge you wish to find. The ROI's rotation bar
should be perpendicular to the edge. When the program is run the
tool will mark the edge it has found.
Edge Polarity
Edge Polarity allows the user to determine whether the Edge Finding
tool will find a dark to light edge or a light to dark edge or simply
the strongest edge present (Don't Care in the menu).
The first image is a light to dark polarity so that the left edge is found. The second
image is a dark to light polarity so that the right edge is found.
Smoothing
Smoothing is the amount of averaging in determining the edge.
Units are pixels.
If smoothing is 0 then a binary edge is sort. Minimum Contrast becomes the threshold. At the threshold the edge is said to be found
If smoothing is -1 then a relative binary edge is sort. Minimum Contrast becomes the min acceptable contrast. At the threshold the edge is = Min + m_Leniency*(Max-Min)/100;
The First image is set to "1"
smoothing, and the second image is set to "20" smoothing.
Leniency
Leniency is a measure of edge steepness. The bigger the number the more
likely the edge finder will find more gradual edges. Units are pixels.
The first image is set to "1"
leniency, and the second image is set to "10" leniency.
Minimum Contrast
The minimum difference in intensity for an edge to be regarded as an
edge. This is measure of greylevel difference. Note that typically for
a greylevel image 0 is black and 255 is white. The 255 would the the
maximum possible contrast.
For the first image the minimum
contrast is set to "15" so it has a hard time finding that
edge. The second image found the edge because the minimum contrast was
reduced to "5".
Show Workings
If checked, the ROI for the different edge points, as these
are found, appears.
ROI Width
The number of pixels / size of the internal sub ROI used to determine
an edge. The larger the number the less sensitive the edge points are
to small edge changes. Units are pixels.
ROI Step
The number of pixels between consecutive edge points. Together with the
ROIWidth this determines the number of edge points found. Units are
pixels.
Which Edge
The Edge of interest. The brightest or the nth edge.
Run:
the model is searched for. Its position and score are calculated
Cycle /Time
reports the execution time in msec of the trained model
Number Edge
The number of points/features found
Each result can be selected. For each result the System calculates:
Contrast
The measure of contrast. !00% then the contrast is the same
as the
trained image
X,Y
The found location of the model