Next: Visual Processing Up: The Dynamic Recursive Previous: How to use

Applying DRFSM in Feature extraction

An experiment was performed for inspecting a mechanical part using a camera and the coordinate measuring machine. A predefined DRFSM state machine was used as the observer agent skeleton. The camera was placed on a stationary tripod at the base of the table so that the part was always in view. The probe could then extend into the field of view and come into contact with the part, as shown in Figure 6.

Symbolic Representation of Features: For the above experiment we were concerned with open regions (O) and closed regions (C). Any closed region may contain other features (the recursive property). Using parenthesis notation the syntax for representing features can be written as follow:

< feature > :: C(< subfeature >) C()

< subfeature > :: < term >, < subfeature > < term >

< term > :: O < feature >

For example, the symbolic notation of Figure 7 is

C(O,C(O,C(),C(O)),C())

Figure 8 shows the graphical representation of this recursive structure which is a tree-like structure. Future modifications to DRFSM's includes allowing different functions for each level.

Figure 9 shows a simple DRFSM DEDS machine for the exploration and inspection of mechanical parts, using both active vision and touch sensors.


sobh@bridgeport.edu
Tue Sep 20 12:46:05 MDT 1994