臺灣博碩士論文加值系統

本研究的目的在於針對3D人體表面資料(3D Body Range data)發展一系列解剖標記點(Anatomic landmarks)標示方法。藉由這些方法可由3D人體掃描資料上擷取出1D、2D、3D的尺寸資料。由於3D人體掃描儀的出現，使得3D人體計測變得更快速而容易，然而要從3D資料上獲得常見的尺寸(參考NASA Webb Associate所整理的項目)是困難的事，主要有三類問題：(1)某些尺寸(如：手臂上舉可及高度、肘角高度)無法直接由標準的解剖姿勢上量測得，而必須擺放特殊的量測姿勢才可量測得到；(2)某些尺寸(如：手指指長、外眼角寬)需要更高的解析的資料才足以辨識(identify)出所需的解剖標記點；(3)某些尺寸(如：顴骨寬度)在傳統上是以觸診(palpation)的方式量測，但在螢幕上要辨識出來不容易。
為了量測出可供平常參考使用且數目足夠的尺寸，本研究設計了4個全身掃描姿勢及7個高解析度的局部掃描姿勢，前者包括3個立姿及1個坐姿，後者包括4種手部姿勢、2個腳部姿勢及1個頭部姿勢。在這些姿勢下標示出120個解剖標記點，然而擷取出231個尺寸，其中191個1D尺寸、26個2D尺寸、7個3D體表面積及7個3D體積。
為了標示出120個解剖標記點，本研究中除了採用文獻中既有的「特徵標示法」及「比例法則」外，另外又提出「局部套用法」及「密度法」以輔助之。「特徵標示法」是由電腦自動標示出表面資料上的幾何特徵點，例如：鼻尖點為鼻子的最凸點、肚臍點為腹部前面的最凹點；「比例法則」是運用到人體的器官的相關位置是固定的道理，例如：鼻尖點是位於頭部約40%~70%頭長高度的位置內；「局部套用法」為將局部的掃描表面資料套用到全身的掃描表面資料上，組合出另外的掃描姿勢，以估計出其他的計測項目的數值；「密度法」是運用單位方格中所包含的三角網格面積的不同，以標示出表面較曲折的部份。其中「局部套用法」及「密度法」為作者的原創作法。

The objective of this study was to develop a research scheme for 3D anthropometrical survey, by designing a series of postures for measurement and then to identify anatomical landmarks based on the 3D measured data for the extraction of 1D, 2D and 3D measurements. Due to advent of 3D human scanner, 3D body anthropometrical measurements become more and more rapid and easier. Nevertheless it was found that it was usually difficult to obtain most of commonly used measurements as listed in Webb Associates. This difficulties can be attributed to three problems: (1) some measurements (e.g. overhead reach or armpit high) can not be taken with a sole standard anatomical posture, rather they should only be taken with specific postures; (2) some measurements (e.g. finger length and outer-ocular) require a higher resolvable data for identification of necessary landmarks in detail; and (3) some measurements (e.g. bi-zygomatic arch breadth) which were traditionally taken by palpation are not easy to identify on monitors.
In order to obtain as many measurements as commonly necessary, this study designed a measurement posture scheme which consists of 4 whole body postures and 7 high resolution localized postures. The former consists of three standing postures and one sitting posture; and the latter consists of four hand postures, two foot postures and one head posture. These postures can be used to identify 120 anatomical landmarks and in turn to extract 231 measurements, 191 1Ds, 26 2Ds, 7 3D surfaces and 7 3D volumes.
In order to identify these 120 anatomical landmarks, segmental concatenation method and density method were proposed in addition to the existing feature-identification method and ratio method. The feature-identification method is used to identify the landmarks with prominent features, for instance, nose tip is the peak in the mid region of mid-sagittal contour. The ratio method, based on statistics, offer the values for the boundary of a landmark, for instance, the nose tip is normally between 40% to 70% of head length. The segmental concatenation method is to concatenate high resolvable local measurements to whole body measurements for specific measurement extraction. The density method is used to identify the landmarks consisting of sharp curvature, for instance, umbilicus is the convex point where the abdominal surface curves in sharply. In the neighborhood of umbilicus, a lattice grid is drawn, for each grid, the number of triangle meshes is counted, all the grids with dense meshes is then picked and then fitted with a curve.

摘要 1
第一章 緒論 4
1.1 研究動機 4
1.2 研究目的 5
第二章 文獻回顧 6
2.1 傳統1D量測儀器 6
2.1.1 馬丁式量測器 6
2.1.2 柵格系統量測器 7
2.1.3 三維量測儀 8
2.1.4 探針式量身器 9
2.1.5 特性比較 9
2.2 3D掃描儀 10
2.2.1 原理 10
2.3 3D人體計測調查 13
2.3.1 日本3D人體計測調查 13
2.3.2 CAESAR計劃 14
2.3.3 比較 15
2.4 解剖標記點的標示方法回顧 16
2.4.1 掃描前貼記號 17
2.4.2 掃描前不貼記號 18
特徵標示法 19
樣版模型法 19
2.4.4 標示法優缺點比較 21
第三章 研究方法 22
3.1 掃描姿勢 22
3.2 局部掃描資料的套用法 36
3.3 掃描儀器及處理軟體 39
3.4 掃描衣著 41
3.5 解剖標記點的規劃及其標示法 42
3.6 標記解剖標記點的法則及演算法 47
第四章 研究結果 50
4.1 解剖標記點的定義 50
4.2 量測項目的定義 56
第五章 討論與建議 71
參考文獻 74

1. 工研院光電所，(2002).「3D技術網頁」，http://3d.itri.org.tw/。
2. 日本人間生活工學研究中心，http://www.hql.or.jp/
3. 林國照，(2001).「數位人體模型資料分析與尺寸量測」，國立清華大學工業工程與工程管理研究所，碩士論文。
4. 洪國力，(2001).「以三度空間人體資料建立尺碼系統暨標準製衣人台」，國立清華大學工業工程與工程管理研究所，碩士論文。
5. 許勝雄，彭游，吳水丕，(1993).「人因工程學」，揚智出版社。
6. 楊宜學，(1998).「三維人體計測-國人頭顱顏面資料庫之建立與應用」。
7. Apparel Research Network , http://arn.iitri.org/
8. Dekker ,L. , I.Douros , B.F. Buxton , P. Treveaven , (1999).”Building Symbolic Information for 3D Homan Body Modeling from Range Data”. Proceedings. Second International Conference on , 1999.
9. Dennis Burnsides,(2001).”3-D Landmark Detection and Identification in the CAESAR Project”, IEEE 2001.
10. Geisen G, Mason GP , Houston V, Whitestone J, McQuiston B, Beattie A,(1995). “Automated Detection, Identification, Registration of Anatomical Landmarks”, Porc. 39th Annual Meeting of the Human factors and Ergonomics Soc. (2), The Society , San Diego,1995,pp.750-753.
11. Hein A.M.Daanen,G. Jeroen van de Water,(1998).“Whole body scanners”，Displays19,pp.111-120.
12. Hertzberg, H.T., Dupertius, C.W., and Emanual , I. , (1957). “Stereophotographmmetry as an anthropometric tool”. Photogrammetric Engineering ,Vol 23, pp. 942-947.
13. ISO 8559: Garment Construction and Anthropometric Surveys —Body Dimensions, International Organization for Standardization, 1989.
14. Mattthcw A. Brunsman, Hein M. Daanen, Kathleen M. Robinette,(1997). “Optimal Postures and Positioning for Human Body Scanning”, IEEE 1997.
15. NASA,(1978),”Anthropometry Source Book”.
16. Nurre JH,(1997).”Locating Landmarks on Human Body Scan Data”,1997 IEEE.
17. The CAESAR project：A 3-D surface anthropometry survey,(1998).
18. Roebuck J.A. , Kroemer K.H.E., Thomson W.G. ,(1975).“Engineering Anthropometry Methods” ,Wiley Series in Human Factors .
19. ROEBUCK J. A.,(1995). “Anthropometric Methods: Designing to Fit the Human Body”, Human Factors and Ergonomics Society, Santa Monaica.