2006 Fiscal Year Final Research Report Summary
Study on an Autonomous Task System Using R F ID
| Project/Area Number |
17560222
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent mechanics/Mechanical systems
|
|---|
| Research Institution | The University of Electro-Communications |
|---|
Principal Investigator |
TAKASE Kunikatsu The University of Electro-Communications, Graduate School of Information Systems, Professor, 大学院情報システム学研究科, 教授 (10270927)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
SONGMIN Jia The University of Electro-Communications, Graduate School of Information Systems, Assistant Professor, 大学院情報システム学研究科, 助手 (00345430)
|
|---|
| Project Period (FY) |
2005 – 2006
|
| Keywords | RFID / Localization / Bayes' theory / Environment Recognition |
|---|
| Research Abstract |
Research results are summarized in the following four items.
(1)Realization of the guidance of a mobile robot using RFID
Localization based on an odometry was not practical because its accumulated error was not negligible. A method of robot pose estimation was proposed, where a tag reader communicated with a number of ID tags which were attached on a ceiling and identified them, then calculated robot pose using detection time information. In the experiment the robot pose estimation error was within 10 cm.
(2)Realization of the guidance of a mobile robot using RFID and TV camera
In addition to RFID, TV camera was used to localize a mobile robot more accurately. This scheme realized the localization accuracy of 2 cm. Based on this method an indoor navigation system was developed and robot mobility was verified even in a corner and long straight aisle.
(3)Realization of the environment recognition for a pick-up task
ID tags were attached on objects to recognize them. RFID information was statistically processed on the basis of Bayes' theory and the position of objects were moderately estimated, while identifying the objects. This method was implemented and experiments were carried out to verify the effectiveness.
(4)Investigation of pick-up task
A method that a stereo vision accurately recognized an object which was roughly localized by RFID was proposed. ID tags bore not only identifier number but also properties of the objects such as color, size, shape, etc.. By utilizing these information, image processing was efficiently and effectively performed for the purpose of pick-up task.
|
Research Products
(9 results)
