Reconstruction and Smoothing of Surface Based on Minimum Energy Principle

Uniform rectangular meshes are constructed to cover all given scattered points．Energy functional of thin plate represented as Hermite surface patches with C¹ or C² continuity throughout the whole data field is mini-mized to fit the data points．Such an approach of global optimization is very effective to suppress the influence of noise in input data and reliable derivatives of first order even second order can be calculated．Effects of smoothing parameter on the precision of surface reconstruction are investigated and the applicability of presented method is il-lustrated by two practical examples．