POINT CLOUD TO SURFACE BEST-FITTING

摘要

Traditional dimensional inspection of a component usually requires that the component be put into a fixture system where its position can be related to its nominal CAD model. This is expensive and time-consuming. Based on a least-squares method, this paper puts forward a new point to surface best-fitting method that eliminates the need of a fixture system, The method has been compared to similar methods, and is less subject to local optimization and has higher accuracy. This is demonstrated using two examples at the end of the paper. For CMM data, this paper integrates a probe compensation method into the best-fitting process.