A new method of point clouds extraction from the 3D volume data

The point clouds is an important kind of three-dimensional data expression form. In the 3D point-based rendering and reverse engineering, it acts as an irreplaceable role. In this paper, the volume data was preprocessed and broken down into a series of 2D slices for contours extraction, finally all of the points extracted were inputted into self-organized network for cluster analysis. The feasibility of this method was proved in both theory and practice. The approach may be taken as a useful reference for researchers working on extracting point clouds from volume data.