3D image segmentation supported by a point cloud

Here, we report a novel method of 3D image segmentation, using surface reconstruction from point cloud data and digital information. For this task, apply mathematical model numerical based on the level set algorithm. This solves by application advection equation with curvature term, which gives evolution an initial condition to final state. is done defining advective velocity in as weighted sum distance function edge detector gradients. The shape, represented cloud, computed fast sweeping method. applied presmoothed image. A crucial for efficiency construction simple tagging algorithm, allows us also highly speed up scheme when solving PDEs. discretization, use semi-implicit co-volume part implicit upwind part. was tested representative examples real representing biological microscopic images developing mammalian embryo.