DocumentCode :
3392257
Title :
Tactile sensor pad: shape recognition
Author :
Famularo, D. ; Muraca, P.
Author_Institution :
Calabria Univ., Italy
fYear :
1995
fDate :
27-29 Aug 1995
Firstpage :
338
Lastpage :
342
Abstract :
Most robot tactile sensing research has focused on object shape recognition with tactile sensors. In this paper we examine the model of a tactile pad sensor, assumed to be a finite dimension medium. The model of the tactile pad sensor is obtained using an integral relation between the boundary conditions and the stress tensor. A set of stress sensors is allocated in the tactile pad, the model is discretized and a direct relation between the parameters of the shape to be recognized and the measured stress is obtained. Due to the ill-posedness of the problem, classical least squares estimation and regularized estimation are examined. Numerical simulations in order to evaluate the inversion technique with different shapes of the object, are presented
Keywords :
least squares approximations; pattern recognition; tactile sensors; boundary conditions; classical least-squares estimation; finite dimension medium; ill-posed problem; inversion technique; object shape recognition; regularized estimation; robot tactile sensing; stress sensors; stress tensor; tactile pad sensor; Atherosclerosis; Boundary conditions; Elasticity; Integral equations; Robot sensing systems; Sensor arrays; Shape measurement; Stress measurement; Tactile sensors; Tensile stress;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Control, 1995., Proceedings of the 1995 IEEE International Symposium on
Conference_Location :
Monterey, CA
ISSN :
2158-9860
Print_ISBN :
0-7803-2722-5
Type :
conf
DOI :
10.1109/ISIC.1995.525080
Filename :
525080
Link To Document :
بازگشت