“…Multi-view range image integration can be categorized into four major approaches: point based methods [4,10,18,23,37,40], direct surface merging [12,15,26,31,33,35,36,41], volumetric approaches employing implicit surfaces [1,2,7,8,9,21,22,25,27,28] and model based methods [24,32]. Volumetric diffusion [3] and volumetric graphcuts [30] have been proposed to build watertight 3D models from multi-views and especially focus on filling holes in the models.…”