The parameterization of point clouds based on projection and applying to reconstruction of automobile body

Aiming at the problem associated with reconstructing a fitting and smooth surface using random points cloud in the reverse engineering of a automobile body, An improved algorithm was proposed, the algorithm fix on the borderline using the Hoschek method and utilize the Base Projection method to attain the parameterization of the other random points cloud. A fitting method is given for the reconstruction of automobile body. Based on the feature of random points cloud of automobile body, a math model integrating the proposed algorithm was formed. Compared with the interpolation surface from UG, a much more similar one with the original surface was given using the proposed algorithm, and it is propitious to the after treatment of the surface.