An improved progressive lossless compression algorithm

Nowadays, people want real-time remote observation and manipulation of 3D geometric data. So, it is necessary to research good 3D data representation methods. In this paper, we will propose a lossless geometry compression approach improved from the progressive valence-driven compression algorithm. It samples the model in a feature-sensitive fashion, with attempt to align the sampling vertices, called anchors, with sharp edges or the surface feature contours. Consequently, without increasing the compression ratio, our algorithm´s distortion ratio performance is better than the original and other recent similar algorithms, especially at low bit rates.