1. 1Mar. 19, 2015 @ SS24 2,11:40-12:00 AM, Floor 1 - Room Lebrija, Meliá Sevilla
A Force-Controlled Portrait
Drawing Robot
IEEE International Conference on
Industrial Technology
March 17-19, 2015
Seville, Spain
Shubham Jain, Prashant Gupta, Vikash Kumar
Swami Keshvanand Institute of Technology (SKIT), India
Kamal Sharma
Bhabha Atomic Research Centre, India
4. 4
Introduction
A Robot that draws the portrait of a human face like an artist.
All current portrait drawing systems can draw only over a flat
pre-calibrated surface.
Motivation
Outcomes
Now, robot is equipped with the force sensing capability, that can make the portrait
like a human artist.
A portrait drawing robot is very enjoyable recreational use of robotics and increases
human-robot interaction.
5. 5
Experimental Setup
Image is captured by Wolf Vision EYE – 12 camera .
Uniform background was preferred so that significant features of face can be
extracted.
For this project KUKA KR6 ARC Robot is used.
Good Illumination conditions must be present.
This robot is equipped with ATI 6 Force/Torque Sensor.
7. 7
Image Processing
Captured image is an RGB image.
RGB image is converted into grey scale image.
Grey scale image is sent for noise removal
using Gaussian filter.
B is a filtered image and A is an image input.
Pre Processing
8. 8
Image Processing
Edge Detection
Edges are detected from preprocessed image using Canny edge
detection algorithm.
Good detection: detects as many real edges as possible.
Good localization: Edges marked should be as close as possible to the real
edge.
Minimal response(towards noise): An edge in an image should be
marked only once and where possible.
Why
Canny
Edge
detector?
12. 12
Image Processing
Post Processing
Labeling of Edges:
Edges extracted are in the form of
binary pixels.
Edges represented by binary
pixel ‘1’.
Assigning label to each pixel by 8
connectivity.
13. 13
Image Processing
Post Processing
End Point Detection:
End points of the edges are detected
Removal of one pixel is necessary
to break the cycle
Robot draw the edge from start
point to end point.
Sorting of Edges:
14. 14
Force-Controlled Robot Motion
Need of Calibration is removed by the use of force sensing strategy.
Robot is mounted with a Force/Torque Sensor at its wrist.
Indirect Hybrid Force/Position Control Mechanism is Used.
Apply Force in X Direction.
Move in Y Direction.
15. 15
Force-Controlled Robot Motion
Draw the next edge point tangentially
over drawing surface
Orient drawing pen normal to drawing
surface
Maintain a contact force between drawing
pen & drawing surface
18. 19
Conclusion
A portrait drawing robot is very enjoyable recreational
use of robotics and increases human-robot interaction.
Humans also maintain a constant force while drawing portraits
Now the portrait drawing robot is equipped with the ability
to sense contact forces.
Now the robot can draw the portrait over a non-calibrated flat
surface as well as other surfaces of any shape.
19. 20
References
Jean-Pierre Gazeau, Said Zeghoul, "The artist robot: a robot drawing like
a human artist," IEEE International Conference on Industrial Technology,
pp.486,491, 19-21 March 2012.
P. Tresset and F. Leymarie, “Portrait drawing by Paul the robot”, Computers
and Graphics, pp. 348.363, 2013.
C. Y. Lin, L. W. Chuang, and T. T. Mac, .Human portrait generation system
for robot arm drawing,. Proceedings of the IEEE/ASME International
Conference on Advanced Intelligent Mechatronics, (Singapore),
pp. 1757.1762, IEEE, July 2009.
S. Calinon, J. Epiney, and A. Billard, “A humanoid robot drawing human
portraits,”Proc. IEEE-RAS International Conference on Humanoid Robots,
2005, pp. 161–166.
20. 21
References
M. Ruchanurucks, S. Kudoh, K. Ogawara, T. Shiratori, and K. Ikeuchi, .Humanoid
robot painter: visual perception and high level planning,. Proc. IEEE International
Conference on Robotics & Automation, pp. 3028.3033, 2007.
A. Srikaew, M. E. Cambron, S. Northrup, R. A. Peters II, M. Wilkes, and K.Kawamura
, .Humanoid drawing robot,. IASTED International Conference on Robotics and
Manufacturing, 1998
Canny J.,. A computational approach to edge detection,. IEEE Trans. Pattern
Analysis and Machine Intelligence, 8:679-714 , 1986.
Kamal Sharma, Varsha Shirwalkar, Prabir K. Pal, “An intelligent indirect hybrid
force/position controller for accurate and smooth tracking of unknown contours,”
Proceedings of Conference on Advances in Robotics, 2013.
21. 22
References
L.F. Baptista, S.P. Fernandes e J.M.G. Sá da Costa, "Bi-dimensional control with
force tracking control in non-rigid environments,“Portuguese Control Conference.
Controlo 2004, 7-9 de Junhode 2004, Universidade do Algarve, Portugal.
Kamal Sharma, .Thesis titled developing and optimizing strategies for robotic
peg-in-hole insertion with a force/torque sensor.,
http://www.scribd.com/doc/106529483/Kamal-sThesis
Donald Hearn, M. Pauline Baker Computer Graphics C Version Second Edition,
Prentice Hall.