In recent years, as the increase of people’s focus on the virtual reality technology, how toget a better three-dimensional reconstruction effect causes more researchers’ attention, and inthe three-dimensional reconstruction process, the simplification of the three-dimensional pointcloud data is the most important part of the three-dimensional point cloud data processing, sothe research on point cloud simplification of three-dimensional reconstruction is a researchhotspot. This paper mainly focuses on how to get a three-dimensional point cloud by thethree-dimensional point cloud data of the different viewpoints and the three-dimensionalsimplification after pretreatment. The important parts of this paper are as follows:Firstly, by using the Kinect camera to obtain a three-dimensional point cloud data underthe different viewpoints of the objects, when pretreatment, an improved ICP algorithm ondifferent viewpoints was presented and uses this algorithm for three-dimensional point cloudregistration under the different viewpoints to get the whole point cloud data of the objects.Secondly, an improved algorithm of the simplification of the three-dimensional pointcloud data is described. By using octree for three-dimensional point cloud data topologydivision, and then use the method which combine the curvature and normal vector to get thepoint cloud feature information, according the information to simply the three-dimensionalpoint cloud, which avoiding the excessive simplification of three-dimensional point cloud onthe curved surface, and make better reconstruction results.Finally, for the features of Kinect camera in getting three-dimensional point cloud data,this paper present a simplification method which combine the color image, this method canavoid the feature information loss of three-dimensional point cloud data where the curvatureof the surface no significant change but it is a feature information, so surface featureinformation can be well preserved by using this method. |