由3D CAD模型自動化辨認加工特徵是自動化產生製造資訊的首要工作，過往研究大都透過特徵資料庫來進行加工特徵的匹配，導致對於特徵間的相交及特徵設計變更的辨認能力較差，且特徵資料庫為有限，而幾何變化為無限，因此沒有任何一種辨識方法能應付所有不同造型的特徵。為克服此問題，本論文以深度學習模型進行加工特徵辨識，有別於傳統辨識只有成功及失敗之分，深度學習模型會根據特徵相似度進行辨識，改善辨識相交特徵之準確度。
本論文透過分析3D CAD模型拓樸與幾何資料，以圖形解析的方法自動化將各個特徵的組成面獨立出來，並以這些組成面進行編碼，以該編碼做為輸入資料，進行深度學習神經網路的訓練，接著使用經過訓練的深度學習模型自動化辨識加工特徵的類別，即可將各個加工特徵顯示在3D CAD模型上，以利後續進行自動化製程規劃。
本論文除了詳述如何分類3D CAD模型之拓樸與幾何資訊、使用圖形解析進行凹凸特徵的搜尋、以取得之特徵資訊進行特徵編碼、利用各種加工特徵之變異形狀產生特徵資料集、建構及訓練深度學習模型外，也利用兩個案例驗證所開發系統的可行性與實用性。

Auto-recognition of machining features based on 3D CAD model is the first step in automatic generation of manufacturing information. The methodologies proposed in past research focused on developing feature library for matching of machining features. However, a feature library has limited data and cannot capture unlimited forms of geometric shapes.It led to low recognition ability of intersecting features and design changes. In order to overcome these problems, this thesis approaches the subject by adopting deep learning model to automate the recognition of machining features. Rather than making conclusions of whether geometric shapes are matched or unmatched, deep learning model can indicate the degree of resemblance and identify the features. Improve the recognition accuracy of intersecting features.
In this thesis, through analyzing the topological and geometrical information of the 3D CAD model, the composing surfaces of each feature can be automatically captured. Based on the geometric properties of various surface adjacency, an appropriate encoding scheme is developed to generate a unique code for each feature. Subsequently, these feature codes become inputs into a pre-trained deep learning model and the type of machining features can be determined. All the machining features can also be shown on the 3D CAD model, allowing the subsequent planning of machining processes.
In addition to proposing the methodologies to classify the topological and geometrical information of the 3D CAD model, search for convex and concave features by graph-based method, feature encoding based on surface properties, generate feature dataset using various shapes of machining features, train and establish the deep learning model, this thesis also uses two case studies to verify the feasibility and practicability of the developed system.

[1] Shah, J.J., Anderson, D., Kim, Y.S. and Joshi, S. (2000), “A discourse on geometric feature recognition from CAD models,” Journal of Computing and Information Science in Engineering, Vol. 1, No. 1, pp. 41-51.
[2] Joshi, S., and Chang, T.C. (1988), “Graph-based heuristics for recognition of machined features from a 3D solid model,” Computer -Aided Design, Vol. 20, No. 2, pp. 58-66.
[3] 張書庭，「木工樑柱加工之自動化製程規劃」(2021)，碩士論文，臺灣科技大學機械工程系，台北市。
[4] Babic, B., Nesic., and Miljkovic Z. (2008), “A review of automated feature recognition with rule-based pattern recognition,” Computers in industry, Vol. 59, No. 4, pp. 321-337.
[5] Henderson, M.R. and Anderson, D.C. (1984), “Computer recognition and extraction of form features : A CAD/CAM Link,” Computer in Industry, Vol. 5, No. 4, p.p. 329-339.
[6] 陳宇昇，「木屋樑柱之特徵辨識及特徵加工路徑的產生」(2021)，碩士論文，臺灣科技大學機械工程系，台北市。
[7] Jones, T.J., Reidsema, C. and Smith, A. (2006), “Automated feature recognition system for supporting conceptual engineering design,” International Journal of Knowledge-Based and Intelligent Engineering Systems, Vol. 10, No. 6, pp. 477-492.
[8] Kim, Y.S. (1992), “Recognition of form features using convex decomposition,” Computer-Aided Design, Vol. 24, No. 9, pp. 461-476.
[9] Waco, D.L., and Kim, Y.S. (1994), “Geometric reasoning for machining features using convex decomposition,” Computer -Aided Design, Vol. 26, No. 6, pp. 477-489.
[10] Ding, L. and Matthews, J. (2009), “A contemporary study into the application of neural network techniques employed to automate CAD/CAM integration for die manufacture,” Computers & Industrial Engineering, Vol. 57, No. 4, pp. 1457-1471.
[11] Ding, L. and Yue, Y. (2004), “Novel ANN-based feature recognition incorporating design by features,” Computers in Industry, Vol. 55, No. 2, pp. 197-222.
[12] Sunil, V.B. and Pande, V.B. (2009), “Automatic recognition of machining features using artificial neural networks” International Journal of Advanced Manufacturing Technology, Vol. 41 No. 9, pp. 932-947
[13] Jian, C., Li, M., Qiu, K. and Zhang, M., (2018), “An improved NBA-based STEP design intention feature recognition,” Future Generation Computer Systems, Vol. 88, pp. 357-362.
[14] NX Open (2014), Siemens PLM Software Inc., Plano, TX, USA.