A fuzzy classifier for tactile sensing

In this paper we propose a fuzzy rule-based algorithm for solvingclassification problems related to tactile sensing. A tactile sensing systemis a robotic device which gives an image of contacting objects. While thefine form recognition problem has been widely discussed and severaltechniques have been proposed for its solution (Bayesian approach or Neuralalgorithms), less attention has been paid to the problem of deciding whetherthe object belongs to a particular class or set of objects that share acommon feature, also known as tactile primitive. As input data we considerthe sum of the normal stresses at the sensing sites. Three levels ofclassification, hierarchically connected, are analyzed and, for each level,different basis variables with their membership functions are proposed andcalibrated using a training procedure. The output is an answer regarding thefeatures of the object at each level and is related to the truth values ofthe fuzzy classes. The numerical experiences show that, at least for dataaffected by low noise level, the algorithm has a very high percentage ofcorrect answers.[PUBLICATION ABSTRACT]