| Title |
Air Pressure Sensor-based Tactile Sensing Using Deep Neural Network |
| Authors |
김동언(Dongeon Kim) ; 이종학(Jonghak Lee) ; 장샤오루(Xiaolu Zang) ; 이장명(Jangmyung Lee) |
| DOI |
https://doi.org/10.5573/ieie.2020.57.4.95 |
| Keywords |
Tactile Sensing Module; Deep Neural Network; Air Pressuer Sensor; Compliant Grasping |
| Abstract |
In this paper, a new tactile sensing module has been proposed using air pressure sensors to sense the tactile sense of the robot hand. In order to provide the force sensing capability of the tactile sensing module, the estimation of force sensing has been learned using the three air-pressure sensor data with Deep neural network. AoT (Arrival of Time) algorithm has been adopted for recognizing the contact point of the robot hand with the object using the tactile module. Resolution of the tactile sensor has been extended to 6×4 with artificial neural network, which might be limited to 3×3 without the artificial neural network. The accuracy and effectiveness of the tactile module has been verified by real grasping experiments. |