200818056 九、發明說明： 【發明所屬之技術領域】 本發明係涉及一種擬態模型及其產生方法，特 一種複合曲面體表模型及内部骨架系統之可驅動=』 之創新設計者。 狄L杈型 【先前技術】 、，隨著電腦圖學與資訊科技的進步，動晝模擬已經漸漸 成為產業應用的重點項目，對於數位人體模型的需求於是 隨之與日倶增。 數位人體模型的組成通常包括靜態的屬性（例如人體 計測資訊、表面外觀…等）以及動態的屬性（例如動作模 式、生理性質…等），相關的研究與技術也往往分成這兩 大範可來發展，較缺乏兼具靜態與動態屬性的數位人體模 型。 、 在數位人體模型的靜態屬性發展上，最基本的是利用 人體計測資訊的表示，亦即身高等各種尺寸，例如jack、 、 職1s等軟體，只要用簡單的幾何表#，就可以順利達到 砰估之目的，不過藉此所產生的模型相似度較低，也無形 中降低其„兒服力。因此，為了提高外觀的相似度，三維掃 描儀可說是廣泛被採用的一項工具，因為其擷取的資料具 有高度的外觀描錢力；有些相關研究是|接根據點與點 2關聯性建立三角網格以產生模型，其他則是以重要的點 0料作為控制點，產生較為平滑的曲面模型。不過，無論 是哪種方法，產生的仍只是靜態、不可驅動的模型，因此 5 200818056 能夠延伸的應用也比較有限。 而在數位人體模型的動態屬 〜屬性發展上，相關研究建立 了各種數學模式來模擬人類的杆 幻仃為動作，但其實際應用的 狀況往往只是透過電腦運算所獲得的數值，缺乏視覺化的 呈現’因此總令人感覺有些缺憾。為了彌補這一點，另有 需多研究係以骨架結構表示人辦 _ 傅衣不人體，將模擬與評估的情況透 過視覺化的呈現，增加传用去μ、廿立+ 曰便用者的滿意度；不過骨架畢竟只 是人體内在、不可見的組成部份， 。根據中華民國專利公報登記資料 與實際的情況仍有差距 ’相關技術僅有兩案： 其中’申明案5虎94132645所揭露之「三度空間人體掃描資 料的自動化特徵點辨識方法」亦為本案發明人之創作，惟 該案之資料處理對象為三維架構中的點資料，其點與點之 間並不具有關聯性’亦無法加以驅動，因此本案可視為該 發明之延伸；而另一案申請案號89426534所揭露之「影像 輔助3D曲面重建與曲面積量測系統」僅涉及曲面之建立。 至於其他國家的專利中，英國的專利，，Generating 3D匕办 models from scanned data (GB 2 389 5〇〇 A)” 雖然同樣是利用掃 描資料建立曲面模型，但其模型是靜態的，並無法驅動； 而美國的專利” System and method for characterizing animatable character (US 6, 384, 819 Bl)，，建立了 具有骨架系統 的客製化可驅動模型，但其活動僅限於二度空間的平面上 有鏗於此，發明人本於多年從事相關產品之製造開發 與没计經驗’針對上述之目標，詳加設計與審慎評估後， 終得一確具實用性之本發明。 200818056 【發明内容】 複合曲面體表及骨 主要希望結合數位 技術特點，大致上 似性的可驅動數位 架的末端點達成帶 於動晝模擬與人因 利用具有高度描述 面體表模型，故能 架系統可透過三維 令該内部骨架上的 位置，藉此以提升 本發明之主要目的，係在提供一種 架之可驅動擬態模型及其產生方法，其 :體模型的靜態與動態屬性，本發明之 是利用三維掃描儀建構具有高度外觀相 擬態曲面體表模型，復藉由控制内部骨 動外部曲面體砉握％丨i 聪表杈型之目的，以利應用 工程評估的相關領域。纟中外觀部份可 力的三維掃描資料手段所建立的平滑曲 提高相似性與美觀的程度；而該内部骨 掃描資料獲得'亦可獨立建置而成，並 關節點與肢節末端點均能夠貼近真實的 動作之正確性及擬真性。 【實施方式】 請參閱第1圖所示，係本發明複合曲面體表及骨架之 可驅動擬態模型及其產生方法之較佳實施例，惟此等實施 例僅供說明之用’在專利申請上並不受此結構之限制；本 發明所述「複合曲面體表及骨架之可驅動擬態模型」就構 成要件而言係包括： 一曲面體表模型10，具備預定之虛擬立體型態，且在 此所述之曲面體表模型10並不限於人體型態、也可以是動 物或卡通型態者； 一内部骨架20，得與前述曲面體表模型之型態相配合 ’且該内部骨架20係與該曲面體表模型1〇相結合； 7 200818056 驅動機制，藉以令該曲面體表模型10與内部骨架2〇 產生關連之作動。 本毛明係可利用三維掃描資料之手段建立可驅動的模 型，該模型是由外部的曲面體表模型10與内部骨架2〇系統 組成。根據實施的步驟，如第2圖所示，可以分為「利用 掃描點資料建立曲面體表模型」、「建立内部骨架」、以 及「將曲面體表模型與内部骨架結合並建立驅動機制」， 兹分別說明如下·· 1、 利用掃描點資料建立曲面體表模型： 在此卩自4又主要疋依據點到線、線到面的順序來建立曲 面，如第3圖所示，首先將三維掃描點資料視為建立 NURBS曲線的控制點，並且將相同水平高度的控制點 41依序串連起來，如此可求得逼近體表輪廓的NURBS曲 線42，接著，再利用曲線間對應的關係，構成平滑的 NURBS曲面，於是完成外觀模型43 (即曲面體表模型10 )的建立。 2、 建立内部骨架： 係可藉由特徵點辨識技術，如輪廓分析、最小體表圍 度、灰階值债測、人體等高線等辨識方法（此技術請 參本發明人相關前案申請案號：94132645所揭），找 出會衫響活動的重要關節點21與肢段端點22 (如第工 圖所不），進而將這些點連接起來，形成内部骨架系 統20 ;隨後可採用逆向運動學（1贿心匪心，ικ) 的方法，即可控制模型的動作，例如：當使用者拉動 8 200818056 - 模型的肢段末端點時，相關的關節點就會自然地跟著 移向適合的位置，於是造成其肢段的動作狀態。 3、將曲面體表模型1〇與内部骨架20結合並建立驅動機制 在建構出人體模型的曲面體表模型1〇與内部骨架别之 後，最後的工作就是將它們結合在一起，使得在控制 内部骨架20的同時，即可帶動該外部之曲面體表模型 10 一起產生動作。本發明將曲面體表模型10和内部骨 ί" 架20結合的邏輯是：該曲面體表模型的控制點可跟隨 内部骨架相對應之關節移動，而根據其相對位置與關 係，被影響的程度又有所不同；因此利用這個特性訂 定曲面受各關節影響的權重值，即可達成藉由内部骨 架驅動外部曲面體表模型之目的；此如第4圖所示， 每一段内部骨架20的控制範圍可以用内、外層封套31 、32來界定，在内層封套31之内的曲面完全不受影響 而在外層封套31之内的區域將可直接跟隨内部骨架 , 20移動，至於介於該内、外層封套31、32之間的區域 (如該苐4圖之Al、Α2所指部位），則可平滑地形變 以因應不同段骨架之間的移動變化，藉此以達到模擬 真實人體肌肉受到拉引時之變化狀態，如此，便能藉 由拉扯該内σ卩骨架2〇來驅動曲面體表模型，以模擬 出真實自然的動作型態，如該第4圖中内部骨架2〇之 關節點21左方之區段產生向上之肘動時，該靠近關節 點21之内、外層封套31、32間的上方區域A1將產生鬆 弛現象（如箭號Li所示），相對地，該靠洱關節點灯 9 200818056 • 之内、外層封套31、32間的下方區域A2將產生拉緊現 象（如箭號L2所示），藉此而能達到擬真之效果者。 最後可將本發明所揭之方法整合於電腦動畫軟體之 中即可驅動利用二維掃描資料建立的客製化模型，以達 成各種動作的模擬。經由將真實的動作與利用本方法驅動 之動作疊合進行驗證，發現兩者具有高度的相似性，另外 ，比較使用本方法產生的模型與真實狀況中關節點的位置 與肢段的長度，兩者之間僅存在微小而可以接受的差異， Q此無娜透過主觀或客觀的方式，都得以驗證本發明具 有實用且可靠的功效。 本發明可應用之範疇如下： 1、 二維人體掃描儀相關軟硬體產業： 搭配二維人體掃描儀的使用，本發明可以拓展其功能 ，除了產生具有高度相似性的外觀，也可藉由控制内 部骨架，驅動其整體模型完成各種動作，得以有效地 提升應用軟體之功能性與吸引力。 2、 產品設計· 利用本發明所產生的可驅動模型，除了可以靜態地測 試產品的合身性之外，也能藉由動態的模擬實現更多 的評估，另如與虛擬的服飾結合，便可測試服裝的活 動性。 3、 工作站設計： 對於製造業而言，當需要設計一個新的工作站時，可 以完完全全地在虛擬的環境中完成評估，包括物件的 10 200818056 配置以及盘A s u ^ 4 - ^ 貝的互動、工作動線的安排，將有助於 大幅降低成本舆人力。 4、 娛樂產業： 纟論是電影、電視或是電玩的製作，都愈來愈依賴電 ::動畫的輔助’藉由本發明得以產生客製化的可驅動 模型，讓每—個角色、玩家都能實現倘徉在虛擬世界 中的新鮮感。 【本發明之優點】 本發明主要是利用三維掃描儀建構具有高度外觀相似 性的可驅動數錢態#面體表模型，復藉由㈣内部骨竿 的末端點達成帶動外部曲面體表模型之㈣，藉此而能完 吴結合數位人體模型之靜態與動態心，以利應用於動書 模擬與人因工程評估的相關領域q中外觀部份可利用: “度描述力的三維掃描資料手段所建立的平滑曲面體表 棋型’故能提高相似性與美觀的程度；而該内部骨架*** 亦:透過三維掃描資料獲得，令該内部骨架上的關節點舆 肢節末端點均能夠貼近真實的位置，藉此以提升 確性及擬真性。 上述實施例所揭示者係藉 雖透過特定的術語進行說明， 利範圍，熟悉此項技術領域之 神與原則後對其進行變更與修 等變更與修改，皆應涵蓋於如 定範疇中。 以具體說明本發明，且文中 冨不月色以此限定本發明之專 人士當可在瞭解本發明之精 改而達到等效之目的，而此 后所述之申請專利範圍所界 11 200818056 【圖式簡單說明】 思圖。 法文字方塊圖 體表模型之示 、外兩層封套 第1圖·係本發明之可驅動擬態模型構成示 第2圖·係本發明之可驅動擬態模型產生方 〇 第3圖·係本發明利用掃描點資料建立曲面 意圖。 第4圖：係本爹明之内部骨架控制範圍用内 界定之示意圖。 【主要元件符號說明】 10 曲面體表模型 20 内部骨架 21 關節點 22 肢段端點 31 内層封套 32 外層封套 41 控制點 42 曲線 43 外觀模型 12200818056 IX. Description of the invention: [Technical field to which the invention pertains] The present invention relates to a mimetic model and a method for generating the same, and an innovative designer of a composite curved surface model and an internal skeleton system that can be driven. Di L杈 type [Prior Art] With the advancement of computer graphics and information technology, dynamic simulation has gradually become a key project for industrial applications, and the demand for digital human models has increased. The composition of a digital human body model usually includes static attributes (such as human body measurement information, surface appearance, etc.) and dynamic attributes (such as action patterns, physiological properties, etc.), and related research and techniques are often divided into two major models. Development, the lack of digital human body models with both static and dynamic attributes. In the development of the static properties of the digital human body model, the most basic is to use the representation of the human body measurement information, that is, the height and other various dimensions, such as jack, job 1s and other software, as long as the simple geometry table #, can be successfully achieved The purpose of the evaluation, but the resulting similarity of the model is low, and it virtually reduces its ability. Therefore, in order to improve the similarity of appearance, the 3D scanner can be said to be a widely used tool. Because the data it captures has a high degree of appearance; some related research is to create a triangle mesh based on the correlation between point and point 2 to generate the model, and the other is to use the important point material as the control point. Smooth surface model. However, no matter which method, it still produces a static, non-driven model, so the application that can be extended in 200818056 is limited. In the dynamic genus-property development of digital human body model, related research Various mathematical models have been established to simulate the human illusion of action, but the actual application is often through computer computing The value obtained lacks visual presentation's. Therefore, it always feels a little lacking. In order to make up for this, there is a need for more research to express people with skeleton structure. 傅傅衣不身 human body, visualizing simulation and evaluation The presentation increases the satisfaction of the users who use the μ, the erection + the squat; but the skeleton is only a part of the human body that is invisible, and there is still a gap between the registration information of the Republic of China Patent Gazette and the actual situation. There are only two cases related to the related technology: The "Automatic Feature Point Identification Method for Three-Dimensional Space Scanning Data" disclosed in the 'Declaration Case 5 Tiger 94132645' is also the creation of the inventor of the case, but the data processing object of the case is in the three-dimensional structure. The point data, the point is not related to the point 'can not be driven, so this case can be regarded as an extension of the invention; and the image-assisted 3D surface reconstruction and curved area disclosed in another application No. 89426534 The measurement system only involves the creation of surfaces. As for patents in other countries, the British patent, Generating 3D models models from scanned data (GB 2 389 5〇〇A). Although the same model is used to build the surface model, the model is static and cannot be driven. And the US patent "System and method for characterizing animatable character (US 6, 384, 819 Bl), established a customizable driveable model with a skeleton system, but its activity is limited to the plane of the second space. Here, the inventor has been engaged in the manufacture and development of related products for many years. In view of the above objectives, after detailed design and careful evaluation, the present invention has finally become practical. 200818056 [Summary of the Invention] The composite curved surface surface and bone mainly hope to combine the digital technical characteristics, and the approximate point of the driveable digital frame can be driven to simulate the dynamic simulation and the human factor utilization has a highly described surface model. The rack system can transmit the position on the inner skeleton through three dimensions, thereby enhancing the main purpose of the present invention, and providing a driveable mimetic model of the rack and a method for generating the same, the static and dynamic properties of the volume model, the present invention The three-dimensional scanner is used to construct a surface model with a high degree of appearance and a mimic surface, and the purpose of controlling the internal bone external surface is to control the relevant field of application engineering evaluation. The smooth curvature created by the powerful three-dimensional scanning data means improves the similarity and aesthetics; and the internal bone scan data can be independently constructed, and the joint points and the end points of the limbs are both Being able to be close to the correctness and immersiveness of real movements. [Embodiment] Please refer to FIG. 1 , which is a preferred embodiment of the driveable mimetic model of the composite curved surface body and the skeleton of the present invention and the method for generating the same, but the embodiments are for illustrative purposes only. The above is not limited by this structure; the "composite curved surface table and the skeleton of the driveable mimetic model" of the present invention comprises: a curved surface model 10 having a predetermined virtual stereotype, and The curved surface model 10 described herein is not limited to the human body type, and may be an animal or a cartoon type; an internal skeleton 20 is matched with the shape of the curved surface model described above and the internal skeleton 20 It is combined with the surface body surface model 1;; 7 200818056 The driving mechanism is to make the surface body surface model 10 and the internal skeleton 2〇 act in connection. The Maoming system can use a three-dimensional scanning method to create a drivable model consisting of an external curved surface model 10 and an internal skeleton 2〇 system. According to the steps of the implementation, as shown in Fig. 2, it can be divided into "using the scanning point data to create a curved surface model", "establishing an internal skeleton", and "combining the curved surface model with the internal skeleton and establishing a driving mechanism", Here are the following explanations: 1. Using the scanning point data to create a curved surface model: Here, the main surface is based on the order of point to line and line to surface. As shown in Fig. 3, first of all The scan point data is regarded as a control point for establishing a NURBS curve, and the control points 41 of the same level are sequentially connected in series, so that the NURBS curve 42 approximating the contour of the body surface can be obtained, and then the corresponding relationship between the curves is utilized. A smooth NURBS surface is formed, and the appearance model 43 (ie, the curved surface model 10) is completed. 2, the establishment of the internal skeleton: by feature point identification technology, such as contour analysis, minimum body surface circumference, gray scale value debt measurement, human body contours and other identification methods (this technology please refer to the inventor related to the case application number : 94142645), to find out the important joint points 21 of the shirting activity and the end points 22 of the limbs (as shown in the figure), and then connect these points to form the internal skeleton system 20; then reverse motion can be used The method of learning (1 bribe, ικ) can control the movement of the model. For example, when the user pulls 8 200818056 - the end point of the limb of the model, the relevant joint points will naturally move to the appropriate The position, thus causing the action state of its limbs. 3. Combine the curved surface model 1〇 with the internal skeleton 20 and establish the driving mechanism. After constructing the curved surface model of the human body model and the internal skeleton, the final work is to combine them so that the control is inside. At the same time as the skeleton 20, the external curved surface model 10 can be driven together to generate an action. The logic of the combination of the curved surface model 10 and the internal bone frame 20 is that the control points of the curved surface model can follow the joint movement of the internal skeleton, and the degree of influence according to the relative position and relationship of the surface. It is different; therefore, using this feature to determine the weight value of the surface affected by each joint, the purpose of driving the external curved surface model by the internal skeleton can be achieved; as shown in Fig. 4, each section of the internal skeleton 20 The control range can be defined by the inner and outer envelopes 31, 32. The curved surface within the inner envelope 31 is completely unaffected and the area within the outer envelope 31 will directly follow the inner skeleton, 20 being moved therein. The area between the outer envelopes 31, 32 (such as the part indicated by Al and Α2 in Fig. 4) can be smoothly deformed to change the movement between the different sections of the skeleton, thereby realizing the simulation of real human muscles. The state of change when pulling, so that the surface model can be driven by pulling the inner σ卩 skeleton 2〇 to simulate a true natural action pattern, as shown in Figure 4 When the segment to the left of the joint point 21 of the skeleton 2〇 generates an upward elbow, the upper region A1 between the outer joint portion 21 and the outer envelopes 31 and 32 will be slack (as indicated by the arrow Li). In contrast, the lower area A2 between the outer cover and the outer envelopes 31 and 32 will be tensioned (as indicated by the arrow L2), thereby achieving the immersive effect. Finally, the method disclosed in the present invention can be integrated into a computer animation software to drive a customized model created by using two-dimensional scanning data to simulate various actions. By verifying the actual action and the action driven by the method, it is found that the two have a high degree of similarity. In addition, the model generated by the method and the position of the joint point and the length of the limb in the real situation are compared. There are only minor and acceptable differences between the two, and Q is able to verify that the present invention has practical and reliable effects through subjective or objective methods. The scope of application of the present invention is as follows: 1. Two-dimensional human body scanner related software and hardware industry: With the use of a two-dimensional human body scanner, the present invention can expand its function, in addition to producing a highly similar appearance, Control the internal skeleton, drive its overall model to complete various actions, and effectively improve the functionality and appeal of the application software. 2, product design · Using the driveable model generated by the present invention, in addition to static testing of the fit of the product, it can also achieve more evaluation through dynamic simulation, and can be combined with virtual clothing. Test the activity of the garment. 3, workstation design: For the manufacturing industry, when you need to design a new workstation, you can completely complete the assessment in the virtual environment, including the object 10 200818056 configuration and disk A su ^ 4 - ^ shell interaction The arrangement of work lines will help to significantly reduce costs and manpower. 4, entertainment industry: Paradox is the production of film, television or video games, are increasingly dependent on electricity:: the aid of animation 'by the invention to produce a customized driveable model, so that every role, player Can achieve a sense of freshness in the virtual world. [Advantages of the Invention] The present invention mainly utilizes a three-dimensional scanner to construct a driveable money state #面面表 model with high similarity of appearance, and by using (4) the end point of the internal skeleton to achieve the external curved surface model. (4) By doing so, it is possible to combine the static and dynamic heart of the digital human body model to facilitate the application of the appearance of the movable book simulation and human factors engineering. q Appearance part of the q: The established smooth curved surface chess form can improve the similarity and aesthetics; and the internal skeleton system is also obtained through three-dimensional scanning data, so that the joint points on the internal skeleton can be close to the real end points. The position of the above is to enhance the accuracy and plausibility. The above embodiments are disclosed by way of specific terminology, and the scope and principles of the technical field are changed and modified. Modifications are intended to be included in the scope of the invention. The invention is specifically described, and The purpose of the modification is to achieve the purpose of equivalence, and the scope of the patent application mentioned later is 11 200818056 [Simple description of the diagram] Thinking. The graphic diagram of the body chart and the outer two layers of the envelope. The present invention is a driveable mimetic model. FIG. 2 is a diagram of a driveable mimetic model of the present invention. FIG. 3 is a view of the present invention using the scan point data to establish a surface intention. FIG. 4: Internal skeleton control of the present invention Schematic diagram of the range defined by the inside. [Main component symbol description] 10 Curved surface model 20 Internal skeleton 21 Off node 22 Limb end point 31 Inner envelope 32 Outer envelope 41 Control point 42 Curve 43 Appearance model 12