Feature-Preserved Geometry Simplification of Triangular Meshes from LiDAR Sensor

Authors:

Abstract
Terrestrial LiDAR sensor plays an essential role in the acquisition of complete three-dimensional data for modeling. Especially the growing demand for detailed models drives the developments in acquisition and processing of terrestrial data. Many automatic simplification algorithms of triangular meshes from terrestrial LiDAR sensor have been developed in order to achieve rapid visualization speed and interactive frame rates. In this paper, a new automatic method is proposed for the simplification of triangular meshes, which preserves visually important regions of the meshes and thus maintains the geometric features of the original meshes at low levels of detail. The proposed algorithm is driven by half-edge collapse operation in a greedy framework and employs a new quadric error metric to measure geometric error. Based on discrete differential geometry, we present an adaptive weight that helps to preserve the visually important vertices and the overall shape of the meshes. Experimental results show that the new method has a good performance for the improvement of geometric feature preservation and the visual fidelity of simplified meshes.