This content is not included in your SAE MOBILUS subscription, or you are not logged in.
Massive Point Cloud Data Sets and Single Point Measurement Acquisition in the Production Floor Environment
ISSN: 0148-7191, e-ISSN: 2688-3627
Published October 16, 2001 by SAE International in United States
Annotation ability available
Event: International Body Engineering Conference & Exhibition and Automotive & Transportation Technology Congress International Body Engineering Conference & Exposition
In typical production lines of automotive manufacturers, body parts are produced every several minutes. Sample parts are measured at certain intervals with a CMM machine at approximately 30 points to verify correct production. The points measured on each sample part are compared with reference points of a golden part or with the CAD model.
This paper presents results achieved with a high-accuracy non-contact 3D measurement system capable of measuring both the full surface of sample parts and surface points (surface point measurement, or SPM), after production, on the production floor. The technology used enables high-speed image acquisition of large data sets together with CMM-like capability of measuring individual points, both accomplished simultaneously. SPM measurement of a car door requires approximately 15 minutes; measurement of a full door surface, including SPM, requires less than 1 hour if operated manually, or alternatively 3 minutes for SPM and approximately 10 minutes for full surface measurement if performed by a robot-based system. The system is capable of measuring shiny, oily and/or dirty parts, which makes it suitable for measurement right on the production floor.
The resulting surfaces and points measured are aligned to vehicle coordinates by the system, and compared to the CAD design. SPMs are compared to their reference points, and reports are exportable. Examples of series of COP/CAD comparisons will be presented, taken on the production floor, illustrating the gradual development of a production defect and its detection with the full surface data. Such defects can immediately be found, when measuring the entire surface of each part, during production (in process). Corrective action can be taken before inaccuracies get out of control.
CitationMizrahi, E., Gershon, R., and Benady, M., "Massive Point Cloud Data Sets and Single Point Measurement Acquisition in the Production Floor Environment," SAE Technical Paper 2001-01-3053, 2001, https://doi.org/10.4271/2001-01-3053.
- Shashua A., “Algebraic functions for recognition”, IEEE Transactions on Pattern Recognition and Machine Intelligence, 17 (8), pp. 779-789, 1995.
- Shashua A. and Navab N., “Relative affine structure: canonical model for 3D from 2D geometry and applications”, IEEE Transactions on Pattern Analysis and Machine Intelligence, 18 (9), pp. 873-883, 1996.
- Hartley R., “Ambiguous configurations for 3-view projective reconstruction”, European Conference on Computer Vision (ECCV), pp. 922-935, June/July 2000.