Back to home | ||
A line feature-based camera tracking method applicable to nuclear power plant environment YAN Weida1, ISHII Hirotake2, SHIMODA Hiroshi3, and IZUMI Masanori4
1. Graduate School of Energy Science, Kyoto University, Kyoto, 606-8501, Japan (yanweida@ei.energy.kyoto-u.ac.jp)
2. Graduate School of Energy Science, Kyoto University, Kyoto, 606-8501, Japan (hirotake@ei.energy.kyoto-u.ac.jp)
3. Graduate School of Energy Science, Kyoto University, Kyoto, 606-8501, Japan (shimoda@energy.kyoto-u.ac.jp)
4 .Fugen Decommissioning Engineering Center, Japan Atomic Energy Agency, Tsuruga, 914-8510, Japan
(izumi.masanori@jaea.go.jp)
Abstract: Augmented reality, which can support the maintenance and decommissioning work of an NPP to improve efficiency and reduce human error, is expected to be pracitally used in an NPP. AR has indispensable tracking technology that estimates the 3D position and orientation of users in real time, but because of the complication of the NPP environment, it is difficult for its pracital use in the large space of an NPP. This study attempt to develop a tracking method for the pracital use in an NPP. Marker tracking is a legacy tracking method, but the preparation work necessary for that method is onerous. Therefore, this study developed and evaluated a natural feature-based camera tracking method that demands less preparation and which is applicable in an NPP environment. This method registers natural features as landmarks. When tracking, the natural features existing in the NPP environment can be registered automatically as landmarks. It is therefore possible to expand the tracking area to cover a wide environment in theory. The evaluation result shows that the proposed tracking method has the possibility to support field work of some kinds in an NPP environment. It is possible to reduce the preparation work necessary for the marker tracking method.
Keyword: maintenance and decommissioning support; augmented reality; tracking; line feature |
||
Click here for viewing full text. | ||