研究生: |
邱靈均 Ling-Chun Chiu |
---|---|
論文名稱: |
應用影像形變於汽車手繪圖之擬真視覺化 Application Morphing Techniques to Realistic Visualization of Car Sketch |
指導教授: |
陳玲鈴
Lin-Lin Chen |
口試委員: |
梁容輝
Rung-Huei Liang 范振能 Jeng-Neng Fan |
學位類別: |
碩士 Master |
系所名稱: |
設計學院 - 設計系 Department of Design |
論文出版年: | 2008 |
畢業學年度: | 96 |
語文別: | 中文 |
論文頁數: | 78 |
中文關鍵詞: | 電腦輔助設計 、汽車設計 、手繪草圖 、形變 、擬真視覺化 |
外文關鍵詞: | CAD, Car Design, Hand-Drawn Sketch, Morphing, Realistic Visualization. |
相關次數: | 點閱:251 下載:1 |
分享至: |
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
隨著數位科技的演進,汽車在概念設計前端不斷的導入數位設計的部份,不僅止於虛擬的二維、三維或是擬真渲染,實體的正向或逆向工程皆以數位科技的方式導入。為了減少設計時程或降低人為的誤差,本研究是虛擬擬真認知的相關研究,欲提供有別以往的新方式,達到一定的效果以作為後續造形評估之用。過去,需要非常專業人員對設計手繪做細部的擬真渲染;也可能是更為費時的縮比油土模型製作;或是透過CAD人員對概念手繪圖建構三維虛擬模型搭配擬真渲染予以呈現。本研究實驗樣本以快速且簡易的製作為基本訴求,並對研究實驗物的結果做初步的優缺點評估。
設計師在許多概念手繪圖之後的執行,往往會考量需要專業人力與漫長時間的投注。因此,許多的概念設計會被要求暫停執行。若只是純手繪影像則會因為不同設計師的設計手繪圖風格差異過大而無法放在一起評估。然而設計手繪圖建構為虛擬三維模型卻會因為二維與三維的詮釋與空間的不合理,使得三維模型與手繪圖有造形上的差異。
本研究提供了有別於三維模型擬真渲染或縮比模型的另一種評估造形方式,可以藉由資料庫透過概念形變,達到簡單並且快速的途徑。試圖達到比純手繪圖的擬真認知高,在轉化的過程中又希望能夠不更改設計原創的造形,對原本的造形設計只提高擬真認知的部份。
實驗結果主要分為兩個方向作為探討:整體的擬真認知度與各特徵對於造形的表達是否明顯。除此之外,各特徵對於造形的表達是以手繪複雜程度在各階段定義 (Rodgers, P A, Green, G and McGown, A, 2000)做初步的實驗與探討。
整體擬真認知的部份,以平均數而言,圖片經過形變影像合成後擬真認知比純手繪影像高;各部特徵的呈現明顯與否部份,以頭燈的合成影像與其他部分的特徵比較有明顯的差異。手繪結合影像形變對於汽車外觀造形而言,需要最終渲染的手繪圖所包含的影像資訊才足夠。在手繪第三階段時,雖然特徵符號與樣式已完整具備,但是面的資訊只由三維平均車的影像資料是不足的,面的轉折與交代必須由粉彩鋪面的方式,僅只線稿的設計呈現出的特徵輪廓與樣式是不夠的。
Following the evolution of digital technology, the concept design stage of automobiles witnessed a continuous introduction of digital design technologies, including not only 2D, 3D virtual modeling, and realistic rendering, but also and forward and reverse engineering of the physical models. In the past, it required highly skilled craftsman to create detailed, realistic rendering from hand drawn sketches. Even with CAD systems, it is still very time consuming for specialists to construct 3-D virtual models, create realistic renderings, or build scaled clay models from hand-drawn concept sketches. The experimental sample appealed basically in fast and simple production and conducted preliminary evaluation of advantages and disadvantages over the experimental item.
To shorten the design schedule and reduce human variations, this research focuses on developing a method for the rapid, realistic visualization of car sketches. The method seeks to achieve visualizations much more realistic than hand-drawn sketches, yet to retain the original design intent during the process of transformation. This objective is achieved by morphing a photorealistic rendering into a set of characteristic lines defined on the concept sketch.
The resulting visualizations are evaluated based on two criteria: perception of overall realism and presentation of various characteristics of the car design. For the perception of overall realism, the level of realism is higher for renderings obtained using the morphing method than hand-drawn images. For the presentation of various characteristics, there are more significant differences for head lamps than other characteristics of the car. To fully describe the shape of a car, hand-drawn sketches not only need to contain characteristic lines, but also shadings that describe convexities and concavities of the surfaces to be sufficient.
1. 町田英明,(2006)。CARSTYLING第170号,東京都新宿区馬場下町、株式会社三栄書房。
2. 町田英明,(2004)。CARSTYLING第170号,東京都新宿区馬場下町、株式会社三栄書房。
3. 王鉅富,(2003)。「造形於形變過程中與情感意象之關係研究-以汽車造形為例」,國立台灣科技大學設計研究所碩士論文。
4. 黃柏松,(2004)。「汽車外觀造形設計的復古表現形式」,國立台灣科技大學設計研究所碩士論文。
5. M. Tovey and S. Poter,(2004). “Sketching, concept development and automotive deign,”Design Studies 24, School of Art and Design, Coventry University, Priory St., Covntry, CV1 5FB, UK
6. Isaac V. Kerlow, (2006). “The Art of Computer Animation and Effects 3rd E,”John Wiley & Sons, Inc.
7. B. Smith and Rowe, (1996), “Compressed Domain Processin of JPEG-encoded Image, ”Real-Time Imaging.