| 摘要: | 逆向工程是解析成品技術的一門科學，實物經過數位化掃描之後，可得到由量測點組成的三角網格模型，最後經過曲線與曲面的處理而產生CAD模型。然而，量測資料由於精度限制、人為失誤、掃描死角或是其它因素，必須對量測資料進行處理，以確保三角網格品質，並從中萃取出所需的點資料，提供CAD模型重建所需的參考，因此本論文對量測網格資料前處理進行研究，並探討其中兩個關鍵問題：網格修復與網格特徵追蹤。 在網格修復方面，本研究提出孔洞的核心修補演算法，依據修補面需求的不同而發展對應的修補演算法，以增加補洞的效率與穩定性，並兼顧修補面的平滑性與接合處的平順，此外，真實掃描資料網格會有各種特殊的狀況與更多的應用需求，因此本研究整理出10種問題網格的形式，並提出對應解決方法，以建立完整的網格修復整合方案。在網格特徵追蹤方面，本研究發展半自動特徵搜尋演算法以取代傳統的全自動特徵搜尋演算法，只需要在特徵上點選一個種子點，演算法即可自動追蹤出所在特徵的路徑，在搜尋過程中，會由種子點為起點往四周搜尋路徑，並定義成本函式估算路徑的好壞，成本函式除了考量特徵之外也考慮路徑的品質與穩定性，路徑擴展則採用雙向多段的搜尋方式，提昇了搜尋的效率。本研究演算法皆透過大量實際掃描資料測試，證明方法的可行性與適應性。 Reverse engineering is a process of reconstructing a 3D virtual model from digitalizating an existing physical part, in which scanned data are acquired by a scanning device. A triangular mesh, composed of scanned data and topology information, is built then. Owning to device limitation and improper operation, some mesh problems might occur frequently and it is hard to obtain key curves and surfaces, used to reconstruct 3D CAD model, from the triangular mesh. Hence, this study aims to solve mesh problems, focusing on two important mesh processes：quality repair and feature tracking. As for quality repair, this study develops a hole-filling method that automatically chooses the specific algorithm depending on the type of the hole. While the hole is small and simple, the efficiency-oriented algorithm is implemented. While the hole is huge and complicate, the quality-oriented algorithm can be implemented to acquire a smooth filled mesh that has good connectivity with the original mesh. Moreover, this study analyses numerous real scanned data and induct ten type mesh problems that affect mesh quality. To solve these mesh problems, this study proposes a comprehensive method. As for feature tracking, this study proposes a new semi-automatic feature detection algorithm using one seed point to provide precise searching for feature points. A search graph, containing nodes and its access relationship, provides the candidate points for the search process. A bi-directional, multi-segment search strategy is then proposed to determine the optimized feature path. The cost function is essentially composed of four terms. The first two terms are employed to track the nodes of similar maximum curvatures and directions of minimum curvature variation, while the last two terms are employed to stabilize the path. In sum, this study demonstrates the feasibility of this approach through a great number of testing trials. |
