Advanced

Recognition of Surfaces Based on Haptic Information

Nakano, Tomohiro (2015)
Department of Automatic Control
Abstract
The main topic and focus in this thesis are surface recognition. In the sensing part, the surface information is collected as feature values. There are research approaches which focus on sensing surface tactile information by using sensors with robots. However, there are some disadvantages of using sensors. From that reason, this thesis proposes getting haptic surface information without sensors at the tip of robots. To this purpose, a disturbance observer is implemented to achieve the robust acceleration control and a reaction force observer is implemented to estimate the friction force along surfaces.

In the recognition part, a pattern recognition method needs to be applied for surface recognition. There are some pattern recognition... (More)
The main topic and focus in this thesis are surface recognition. In the sensing part, the surface information is collected as feature values. There are research approaches which focus on sensing surface tactile information by using sensors with robots. However, there are some disadvantages of using sensors. From that reason, this thesis proposes getting haptic surface information without sensors at the tip of robots. To this purpose, a disturbance observer is implemented to achieve the robust acceleration control and a reaction force observer is implemented to estimate the friction force along surfaces.

In the recognition part, a pattern recognition method needs to be applied for surface recognition. There are some pattern recognition methods, where selforganizing maps (SOM) is one of the solutions and has been investigated. SOM is able to summarize high-dimensional data to low dimension with preserving the topological properties of data. SOM is also suitable for multi-class recognition. Therefore, a surface recognition using SOM is proposed in this thesis. Multi-class surface recognition is achieved by the proposed method. The validity of the proposed method was confirmed through 7 surface recognition experiments. The recognition rate was over 90% for 5 of 7 surfaces in the time domain. (Less)
Please use this url to cite or link to this publication:
author
Nakano, Tomohiro
supervisor
organization
year
type
H3 - Professional qualifications (4 Years - )
subject
ISSN
ISSN 0280-5316
other publication id
ISRN LUTFD2/TFRT--5957--SE
language
English
id
5148590
date added to LUP
2015-03-06 09:37:02
date last changed
2015-03-06 09:37:02
@misc{5148590,
  abstract     = {The main topic and focus in this thesis are surface recognition. In the sensing part, the surface information is collected as feature values. There are research approaches which focus on sensing surface tactile information by using sensors with robots. However, there are some disadvantages of using sensors. From that reason, this thesis proposes getting haptic surface information without sensors at the tip of robots. To this purpose, a disturbance observer is implemented to achieve the robust acceleration control and a reaction force observer is implemented to estimate the friction force along surfaces.

In the recognition part, a pattern recognition method needs to be applied for surface recognition. There are some pattern recognition methods, where selforganizing maps (SOM) is one of the solutions and has been investigated. SOM is able to summarize high-dimensional data to low dimension with preserving the topological properties of data. SOM is also suitable for multi-class recognition. Therefore, a surface recognition using SOM is proposed in this thesis. Multi-class surface recognition is achieved by the proposed method. The validity of the proposed method was confirmed through 7 surface recognition experiments. The recognition rate was over 90% for 5 of 7 surfaces in the time domain.},
  author       = {Nakano, Tomohiro},
  issn         = {ISSN 0280-5316},
  language     = {eng},
  note         = {Student Paper},
  title        = {Recognition of Surfaces Based on Haptic Information},
  year         = {2015},
}