數位條紋投射法量測胸腔外形三維點資料之處理與應用

Abstract

本論文研究漏斗胸病患胸腔外形量測三維點資料之處理技巧及其應用，三維點資料乃由兩台不同視角的相機分別利用數位條紋投射法取得。首先發展結合兩組不同視角點資料之自動對位與整合方法。在兩組點資料之對位時，經由指定兩組不同視角點資料之對應相同特徵點，本研究發展一套有效的演算法以找出最佳之座標系統平移及旋轉矩陣以結合兩組點資料。在資料點之整合時，發展平均混成、線性混成及二次混成等三種方法將兩組點資料整合成一組沒有縫隙及沒有多餘部分之平滑表面。接著發展一套軟體系統，可以進行兩組點資料之顯示、對位、整合與後處理等工作，此軟體系統使用Visual C++程式語言發展而成。最後，將本文發展之系統應用於漏斗胸病人之胸腔外形量測，在置入固定桿件手術之前與之後分別量測病人之胸腔外形，經由比較手術前後之胸腔外形，可評估手術之效果。

This thesis studies the processing techniques and applications of 3D data points in the shape measurement of the ribcage for patients with funnel chest. The 3D data points are acquired by the digital fringe projection method from two different camera views. First, automatic registration and integration methods for combining two point sets of different views are developed. In registration of data points, points of the same feature in two different data sets are specified, and a robust algorithm is developed for finding the best coordinate translation and rotation matrices that combine two point sets. In integrating data points, average blending, linear blending, and quadratic blending methods are developed such that two data sets are integrated into one seamless and non-redundant smooth surface. Then, a software system is developed for data points display, registration, integration, and post processing for the two point sets. The software system is developed using Visual C++ language. Finally, the developed system is applied in the ribcage shape measurement for patients with funnel chest. The ribcage shapes of a patient before and after the surgery for placing a fixation bar inside the chest are measured. The effect of the surgery can be evaluated by comparing the ribcage shapes before and after the surgery.