Feature-preserved simplification of point-sampled models

Abstract

We present an approach for simplifying dense point-sampled models with feature preserving. Here we adopt an iterative algorithm of point-pair contraction with a feature-weighted quadric error metric. Our idea is that the cost of contracting a point-pair is penalized by the surface feature weights. Compared with the iterative simplification algorithm with a feature-unweighted quadric error metric, experimental results illustrate that our approach is more efficient in surface feature preservation.