SKYNET

1
Vision-Guided RobotPosition Control

• SKYNET

Tony Baumgartner Computer Science Jeff Clements
Mechanical Engineer Norman Pond Electrical
Engineer Brock Shepard Mechanical
Engineer Nicholas Vidovich Computer Engineer
Advisors Dr. Juliet Hurtig Dr. J.D. Yoder
May 4, 2004
2
Problem Identification
3
Equipment Provided

• Desktop computer
• Two CCD (Charge-Coupled Device ) cameras
• ABB Robot
• Controller

4
Robot
ABB Articulated IRB 140
5
Controller
ABB - S4CPlus
6
Design Deliverables

• Development Completion Date
• Difference Algorithm 12/19/03
• Image Processing 02/02/04
• Object Recognition 02/02/04
• User Interface 02/20/04
• Robot Communication 03/15/04
• Overall
• Gripper Implementation 02/02/04
• Testing 04/30/04

7
Software Block Diagram
8
Object Recognition Algorithms

• Shape-Based Matching
• Thresholding
• Blob Detection
• Pattern Recognition
• Similarity Measure

9
Object Recognition VisualsNo Object in Scene
10
Object Recognition VisualsReference Points Enter
Scene
11
Object Recognition VisualsObject Enters Scene
12
Object Recognition VisualsShowing Blobs
13
Object Recognition VisualsFinding Blobs
14
Other Object Recognition Visuals
15
Other Object Recognition VisualsShow Lines
Enabled
16
Axis Offset
17
Camera Axis
18
Calculating Image Angles/Distances

• ?x tan-1 ((x2 - x1 ) / (y2 - y1))
• ?y tan-1 ((y3 - y1) / (x3 - x1))
• ?1 p/2 ?x ?y
• pxDistanceBetweenRefs v(y2 - y1)2 (x2 - x1)2
• pxDistanceRefOneToObj v(y3 - y1)2 (x3-x1)2

19
Robot Axis
20
Calculating Robot Angles/Distances

• ?a tan-1((x2 x1) / (y2 y1))
• ?b p/2 ?a ?1
• mmDistanceBetweenRefs v(y2 - y1)2(x2 - x1)2
• mmtopxRatio mmDistanceBetweenRefs /
  pxDistanceBetweenRefs
• mmDistanceRefOneToObj mmtopxRatio
  pxDistanceRefOneToObj

21
Robot Coordinates

• Xr1Obj r cos ?
• Yr1Obj r sin ?b
• XObj x1 Xr1Obj
• YObj y1 Yr1Obj

22
Testing

• Testing included
• A protruding device from TCP
• Measuring stick (millimeter)
• 10 different reference points
• 4 different object positions
• 2 positions per quadrant
• Maximum Error 3 mm
• Average Error lt 2 mm

23
Future Considerations

• Implement 3D
• Improve camera resolution to improve accuracy
• Implement improved object recognition algorithms
• Add mechanical style gripper
• Optimize softwares system resource usage

24
QUESTIONS
?
25
References

• lt1gt ABB Product Specification Sheet (2003)
• lt2gt Lin, C.T., Tsai D.M. (2002) Fast normalized
  cross correlation for defect detection. Machine
  Vision. Yuan-Ze University,1-5.
• lt3gt Phil Baratti Robot Precision (personal
  communication, November 4
• 2003)
• lt4gt Stegar, C.., 2001. Similarity measures for
  occlusion, clutter, and
• illumination invariant object recognition. In
  B. Radig and S. Florczyk(eds), Mustererkennung
  2001, Springer Munchen, pp. 145-154.
• lt5gt Stegar, C., Ulrich, M. 2002 Performance
  Evaluation of 2D Object
• Recognition Techniques. Technical Report
  Technische Universitat Munchen, 1-15.
• lt6gt Robots and Manufacturing Automation, pg.
  220-222.
• lt7gt http//www.prip.tuwien.ac.at/Research/RobotVis
  ion/vs.html