Research On The Reconstruction System Of Tooth Model Based On CT And Optical Scan Data

In computer prosthodontic system,achieving the registration of computed tomography and optical scanning tooth model is related to the subsequent process of tooth arrangement,planting,and fusion.The dental model reconstruction system based on CT and optical scan data can quickly and efficiently obtain a complete and accurate three-dimensional tooth digital model.The system is mainly divided into five modules:First,the initial reconstruction of CT data;the second is the separation of the optical crown and the gums;the third is the registration of the CT data after initial reconstruction with the optical crown;the fourth is the common segmentation of the two models;the fifth is the fusion of the two data models.This paper is mainly responsible for the registration and common segmentation of the two models.In order to achieve better registration of the two different resolution data images and reduce the human-made interference of the traditional manual registration,to improve the defects of automatic registration which lead to low registration accuracy and a large number of calculations,an efficient registration algorithm for computed tomography and optical scanning dental models with different resolutions is proposed.Firstly,a correct scale factor was used to create a temporary point set,and generate spatial zoom scaling parameters for isotropic scaling in the initial registration.Secondly,a stridden sampling algorithm was proposed to select h feature points with information retention of the original data for reducing the amount of data calculation.Lastly,an improved ICP register algorithm with 7 degrees of freedom for accurate registration was proposed in precise registration.The experimental results demonstrate that the proposed method provides faster convergence and smaller registration error than existing methods,and can achieve registration of tooth models of different data types.Aiming at the problems of low time efficiency and difficult segmentation of CT data cutting of teeth,this paper proposes a method of co-segmentation of CT data and optical crown data of teeth.First,we use their similarity and non-similarity to accurately segment the crown.Only the same portion of the crown and the CT dental mold is retained,and the different regions are removed,and a small amount of the gingival portion of the optical crown edge is separated.A single crown model will then be isolated and used to guide the precise cutting of the CT model.The optical crown was cut using a CT slice to map the initial contour on each CT slice,and iterated using the classical Level set method to segment the complete tooth model.Finally,the optical crown is accurately segmented again using the complete CT tooth model,and the gap of the individual teeth is sewn to create a complete optical tooth model.The experimental results show that the proposed algorithm has high segmentation efficiency and high speed.