Abstract: A triangular mesh sampling method based on the curvature and distance is proposed for the large scale scattered data. By this approach, the large scale scattered data can be compressed to a reasonable expected scale, and the triangular mesh generated by the compressed data can clearly describe the details of surface features. Every triangle patch of the generated mesh also has good shape. The experiments show that the method is efficient and easy to apply.

Key words: computer application, surface reconstruction, adaptive compression, scattered data