Example 1: Locate a long hole on a chain saw sword

To separate a long hole from other blob can be a little challenge - with the feature classification features in scorpion this is easy. In the example below six blob are found and returned as blob results.

Feature Classification Setup

To classes are defined

• long - activated
• no hole

Long has one reference of a long hole. The no hole class is used to add references of other blobs of no interest.

A maximum distance of 5 is defined. Automatic weight calculation is selected.

The classification vector consists of the following results:

• Contour area
• Circularity
• Axis Ratio
• Angle

This result vector defines a good vector space.

Results after activating Feature Classification

Only the long hole is accepted and returned as a result and no result validation is required.

Example 2: Iterate all center on a toolbox component

ring = GetTool('VerifyRing')
img = GetImageMatr('Image')
center = GetResultValue('tf3.AllCenters')
for pos in center:
  ring.executeCmd('SET','OBJECT=CENTER;VALUE=((%.4f,%.4f,0))' % (pos[0],pos[1]))
  ring.execute(img)