Laser range scanners allow highly accurate measurement of manufaturing parts such as tubes. However, generic surface reconstruction software does not give the specific information needed for quality assurance and cannot provide correction data for the tube bending process. Our aim is to find a tubular geometry which fits best with the measured data.
- Segmentation of point cloud data
- Parametric registration of geometric primitives
- Real-time processing of sampling data
While measuring data from tactile coordinate measurement can be processed by simple nonlinear fitting techniques, differential geometric methods are needed to fully exploit the additional information provided by a laser scanner, such as deformation of the tube. We provide methods for accurate reconstruction of tubular sufaces and provide relevant information for quality assurance and reverse engineering.