圖片中之角色動畫化

Animating Characters in Pictures

戴士強(Shih-Chiang Dai)

指導教授 ： 陳炳宇

國立臺灣大學/管理學院/資訊管理學研究所/碩士(2008年)

https://doi.org/10.6342/NTU.2008.03299

摘要

由一張靜態圖片產生出三維空間動畫，對於製作電影、卡通、動畫特效來說，會
是非常省時省力，又能提高產品品質的方法。本系統提供一介面讓使用者可以藉
由簡單流程，讀取圖片與三維動作資訊後，產生讓圖片中的人物，在三維空間中
作各種動作的動畫。
核心概念是用一個帶有骨架的人物模型作為我們的樣板，圖片中之人型角色由
使用者指定，並指定樣板特徵點與圖片中之人物特徵點的對應關係後，我們以此
為已知條件，利用維持堅固性的演算法，將此樣板三維模型，先在二維空間中把
模型調適至與二維圖片中之人物輪廓一致，藉由此結果拉動關節位置至圖片人物
中之適當位置，最後再根據輪廓與關節的距離，適當調整第三維資訊- 厚度- 至符
合一般人體比例，真正完整重建出一個二維人物在三維空間中的資訊。
比起直接由圖片重建出三維模型的方法，本研究之特點為，在產生三維模型的
同時，輪廓特徵能與二維人物一致，又不會喪失原始模型樣板所具有的正常人體
比例特徵。若要應用至特殊型態的角色，如玩偶，動物，我們只要改變所使用的
樣板，依然能套用本系統之演算法進行，達到使圖片中之腳色動起來的效果。

關鍵字

角色動畫 ； 形變 ； 基於圖片之幾何建模系統 ； 使用者輔助

並列摘要

To generate a 3D animation from a single image, is an effective and economic way in
movie, animation, and game industry. In this paper, we present a system that allows the
user to animate a character in a picture in 3D space by applying an existed 3D character
model with motion data.
The 3D character model with skeleton rigged is used as a template model to fit the
silhouette of the character in the picture. , which is cut out by the user. After assigning
some corresponding points between the 2D character image and 3D template character
model, the system first fits the template to the 2D character silhouette in 2D space, using
an as-rigid-as-possible algorithm. Then the system will compute the new position of
skeleton according to the fitted model. the system adjust the thickness by the distance of
silhouette and bones with a suitable ratio. Then we can transfer the colors and patterns
of the image to the model as the textures. Finally, the user can apply any motion data to
animate the fitted 3D character model in 3D space.
The feature of our work, different from the way that reconstruct the 3D model directly
from the image, is that we can preserve geometry feature of our template, and satisfying
the 2D silhouette constrain at the same time. If we want to fit the character to some
different types of characters, such as dolls, animals, we can simply replace the template.
The system can still apply the same algorithm to do the work. Thus users can achieve
their purpose - animating characters.

並列關鍵字

character animation ； shape deformation ； image-based modeling ； user-aid

參考文獻

[1] I. Baran and J. Popovi´c. Automatic rigging and animation of 3d characters. ACM Transactions

[2] B.-Y. Chen, Y.Ono, and T.Nishita. Character animation creation using hand-drawn sketches.

pictures with stochastic motion textures. ACMTransactions on Graphics, 24(3):853–

[4] I. Drori, D. Cohen-Or, and H. Yeshurun. Fragment-based image completion. ACM Transactions

[5] A. A. Efros and T. K. Leung. Texture synthesis by non-parametric sampling. In Proceedings