Point Cloud Simplification Based On Bounding Box And Maximum Information Similarity Error Measure

With the continuous development of 3D measurement and scanning hardware equipment and the emergence of some new measurement technology,people increasingly high requirements for 3D geometric modeling technology,due to the large amount of data for subsequent processing caused great difficulties,therefore,under the condition of keeping enough key geometry characteristic information needed for the follow-up processing of the measurement object,how to simplify the point cloud data to the maximum extent is very important for accurate,rapid and efficient processing becomes.This paper breaks through the commonly used method of point cloud simplification,and use the method of bounding box and the quadric error metric of the similarity of information.It can be based on the maximum point cloud reduction to better maintain the characteristics of the object.Firstly,aiming at the traditional point cloud simplification method,this paper based on the analysis of the problems of low efficiency and poor efficiency,the bounding box using the point cloud block greatly improving the point cloud search time.Secondly,The information similarity model was established to quantitativly describe the differential geometrical similarity between the sampled points.Based on the principle of maximum information similarity to establish each group of contraction points and information similarity quadric error metric.The optimal position of the contracted pair was solved by minmizing the quadric error matric.The similarity of information provides a guarantee for the accurate and rapid reconstruction of model characteristic curve.Thirdly,the contraction point breakthrough to contract to a point idea,the method of information similarity is used to combine three points into a set of efficient points in maintaining object features that improves the efficiency of contraction.Finally,the experimental simulation is carried out by the bounding box method and the algorithm proposed in this paper,the validity of the method is proved.The point cloud data obtained by 3D scanner,using the method of sampling error,the average error and the maximum error of the simplified model are compared and the simplified point cloudreconstruction in reduction rate,It proves that this paper algorithm the advantages of the feature preserving prominently.