Detection and Inspection Planning for Ship Propeller Blades via Spectral Shape Analysis
| Autor: | Olav Egeland, Eirik B. Njaastad |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
0209 industrial biotechnology
Spectral shape analysis Computer science Acoustics 020208 electrical & electronic engineering ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Propeller 02 engineering and technology engineering.material 020901 industrial engineering & automation Control and Systems Engineering 0202 electrical engineering electronic engineering information engineering engineering Robot Bronze |
| Zdroj: | IFAC-PapersOnLine. 52:154-159 |
| ISSN: | 2405-8963 |
| DOI: | 10.1016/j.ifacol.2019.10.038 |
| Popis: | Inline quality control for one-off or few off manufacturing requires highly accurate and automated 3D measurement systems. This paper presents an approach to the planning of robotic inspection of geometric tolerances for large ship propeller blades cast in NiAl bronze. The spectral shape of the blade is analyzed and compared to a generic propeller blade model so that the blade’s position and orientation in the scene can be determined. The shape analysis relies on the eigenspectrum of a Laplace-Beltrami operator for segmenting the scene and identifying the propeller blade. To avoid time-consuming inspection of the entire propeller blade, the foil sections of the blade are extracted from an initial coarse scanning of the scene. Then the identified foil sections are the basis for generating inspection view poses for a robot inspecting the propeller blade with a high accuracy 3D camera. The system is successfully validated in simulations and on experimental data. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|