應用多視角三維點雲對位於齒模重建之研究;Application of Multi-Angle Registration to Digital Dental Cast

National Chung Cheng University Institutional Repository > 管理學院 > 資訊管理系研究所 > 學位論文 > Item A095B0000Q/494

Please use this identifier to cite or link to this item: http://ccur.lib.ccu.edu.tw/handle/A095B0000Q/494

Title:	應用多視角三維點雲對位於齒模重建之研究;Application of Multi-Angle Registration to Digital Dental Cast
Authors:	林劭儒;LIN, SHAO-RU
Contributors:	資訊管理系研究所
Keywords:	數位齒模;疊代最近點演算法;移動最小平方;three-dimensional image;dental cast;Iterative closest point;point cloud;ICP;Triangular Meshes
Date:	2017
Issue Date:	2019-07-17
Publisher:	資訊管理系研究所
Abstract:	數位牙醫，是以科技輔助牙醫為患者進行矯正，藉由三維影像處理技術，經過掃描、點資料對位、點資料建模，將石膏齒模數位化重建，能夠簡化假牙的製作過程，以及幫助牙醫師做進一步的診斷，實現數位牙醫的目標。在本研究中，提出一系列藉由石膏齒模取得數位齒模的方法，利用投影結構光搭配CCD，針對不同視角的石膏牙齒模型掃描多筆表面數據，根據掃描出的點資料空間關係，辨別出主要的齒模部份並濾除其餘雜訊，接著對影像進行簡化加速其執行效率，以疊代最近點演算法(ICP)計算點資料間的轉換關係，將多筆點雲做對位，使用移動最小平方法整合因重複疊合而產生誤差的點資訊，最後經由Poisson重建出表面的網格，將無序的抽象點數據建構為實體模型，精準的將實體齒模以數位化的方式呈現。同時比較了不同掃描角度之間的差異，透過本研究的方法，能夠以少次的掃描，建構出完整的數位齒模。 In this study, we used the techniques of three-dimensional image scanner and image registration to display the general dental cast in the way of　digitizing. And we focused on image registration in this case. At first, we utilized Statistical Analysis Techniques to eliminate noise from the sets of multi-view dental cast of point cloud data. Then The image registration region we adopted ICP (Iterative closest point) algorithm to calculate the correspondent with the sets of point cloud. Then we overlapped two adjacent dental cast images from a series of image sequence several times to get the complete image information. Finally, we visualized the disorderly sets of abstract-pointed data through the Triangular Meshes. The purpose of the proposed method want to increase the accuracy and availability when dentists run diagnostics on the teeth. Therefore, dentists would be easier to communicate with the patients for solving the problems.
Appears in Collections:	[資訊管理系研究所] 學位論文

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	1075	View/Open

Loading...