Showing The Complete Debug Views, Using CML:
Activating The Gesture Analytics Display, Part 3
Every touch object rendered on stage has the "built-in" ability to
independently display the touch point clusters it posses as well as
detailed informatics on cluster geometry and motion. To turn on the GAD
display, in the "my_application.cml" document set the attribute "displayOn" equal to true. For example:
<DebugKit displayOn="true">
This will globally activate the debug display of all touch objects
rendered on stage. When the GAD is activated it is always rendered on
the topmost layer of the application display hierarchy. This is done to
ensure that, when in debug mode, touch points and touch/gesture data is
always visible and not obscured by media or menus.
When all "DebugLayers" in the GAD are activated, it can show a
visualization of the full gesture pipeline from touch point tracking,
cluster analysis (geometry and motion), gesture processing (noise
filtering, inertia) to property mapping and display object
transformations. In this example we are concerned with the display of
touch point cluster geometry. For example:
<DebugKit displayOn="true">
<DebugLayer type="point_shapes" displayOn="true"/>
<DebugLayer type="cluster_circle" displayOn="true"/>
<DebugLayer type="cluster_box" displayOn="true"/>
<DebugLayer type="cluster_center" displayOn="true"/>
<DebugLayer type="cluster_web" displayOn="true"/>
<DebugLayer type="cluster_rotation" displayOn="true"/>
</DebugKit>
This will activate layers that will dynamically drag the touch point
positions using "point_shapes", define a cluster radius using
"cluster_circle", define the width and height of the cluster using
"cluster box", show the center* of the
cluster using "cluster_center", connect all the touch point to the
center point using "cluster_web" and show the change in rotation in the
cluster using "cluster_rotation".
* There are multiple different types of
cluster centers used in gesture processing. This display shows the
center of the circle defined by the extent of the touch point cluster.
This measure of center gives a good approximation of the palm of a human
hand when using five touch points and is geometrically consistent with
clusters containing other numbers of touch points.
