In this paper, the design and experimental evaluation of a sensor platform, including cameras, intended for inspections in remote or difficult-to-access locations; e.g. offshore wind turbine nacelles are presented. The system consists of a sensor system attached to a car that moves on a rail. The main requirements for the design include applicability for high-volume production of rail segments as well as high-accuracy position of the car and the direction sensors face. This accuracy is important for the use of image change detection algorithms. A series of experiments was conducted and the results show that the position errors are just large enough to have an effect on images taken of components one meter away with a 60-degree field of view 1080p camera. However, it is considered realistic that computer vision algorithm can compensate for this.
展开▼
