当前位置:主页 > 文艺论文 > 广告艺术论文 >

二维卡通动画动作捕捉与重定向

发布时间:2018-06-27 09:21

  本文选题:二维卡通 + 动作捕捉 ; 参考:《天津大学》2012年硕士论文


【摘要】:二维卡通动画因其独特的夸张手法和简单鲜明的表现力深受人们喜爱。传统动画中的卡通角色非常富有表现力,动画师可以通过控制角色的各个参数赋予它们显著的个性和丰富的动作。传统二维卡通动画的制作流程极其繁琐,分工细,每一环节都有专业的人员来完成,所以动画最终才能展现出超出观众想像的表现魅力,能最大程度的模拟甚至超越真人的表演,给观众带来无法替代的、独特的娱乐享受。 但是二维卡通动画制作起来非常麻烦,特别是当制作一些商业短片,广告,,或演示动画时,这些动画的动作重复性大,却依然需要逐帧制作,调整,导致动画制作成本高昂。动画制作者不得不花费大量的时间制作动画中的关键帧,这是一项非常繁琐复杂的工作。大大减少了动画创作者创作的时间,降低了工作效率。 动画重用是指在已有的动画素材基础上,捕捉到源动画的动作或者角色细节,映射到新的动画角色上。所以只要能有效地对已有的卡通动画素材进行重用,就可以很好的解决这一问题,不但可以大大的降低动画制作的成本和制作周期,同时为艺术家们创作节省相当可观的工作量,使他们能专注于艺术创作,充分发挥他们的灵感和原创性。提高国产动画发展水平。 本论文的主要内容如下: 1.从源卡通动画中截取关键帧为其动作捕捉做预处理。利用已有的二维卡通动画,从中自动截取某段包含可以重用动作的关键帧,然后进行背景减除,消除噪声,边缘平滑等预处理。 2.捕捉预处理后的角色动作信息。利用形状上下文捕捉角色的边缘轮廓点和关键点位移坐标,并导出关键点的位移坐标数据。 3.动作重定向预处理。先对目标动画进行提取轮廓点,图形内部离散采样,三角网格化等预处理,再用捕捉到的关键点位移作为控制点驱动角色形变。 4.利用双步闭合形状算法分两步对目标动画进行形变计算,让角色在关键点的控制下做出适当的形变产生对应动作,实现动作重定向。
[Abstract]:Two-dimensional cartoon animation is loved by people because of its unique exaggeration and simple expression. Cartoon characters in traditional animation are very expressive. Animators can give them remarkable personality and rich actions by controlling the parameters of their characters. The production process of traditional two-dimensional cartoon animation is extremely complicated, the division of labor is fine, each link has the specialized personnel to complete, therefore the animation finally can display beyond the audience imagination performance charm, Can maximize the simulation or even beyond the real-life performance, to the audience to bring irreplaceable, unique entertainment. But 2D cartoon animation is very troublesome, especially when making some commercial short film, advertisement, or demonstration animation, the action of these animation is very repetitive, but still need to make frame by frame, adjust, lead to the high cost of animation production. Animators have to spend a lot of time making key frames in animation, which is a very complicated task. Greatly reduces the animation creator's creation time, reduces the work efficiency. Animation reuse refers to capturing the action or character details of the source animation based on the existing animation material and mapping to the new animation character. So long as we can effectively reuse the existing cartoon materials, we can solve this problem very well, not only can greatly reduce the cost of animation production and production cycle, At the same time, artists save considerable amount of work, so that they can concentrate on artistic creation and give full play to their inspiration and originality. Improve the development level of domestic animation. The main contents of this thesis are as follows: 1. The key frame is extracted from the source cartoon animation to preprocess its action capture. The existing 2D cartoon animation is used to automatically intercept a certain key frame which can be reused, then background subtraction, noise elimination, edge smoothing and other preprocessing are carried out. 2. Capture pre-processed role action information. The shape context is used to capture the contour points and the displacement coordinates of the key points, and the displacement coordinate data of the key points are derived. Action redirection preprocessing. The object animation is preprocessed by extracting contour points, discrete sampling and triangulation, and then using the captured key point displacement as the control point to drive the role deformation. 4. The two-step closed shape algorithm is used to calculate the deformation of the target animation in two steps so that the characters can make appropriate deformation corresponding actions under the control of the key points and realize the action redirection.
【学位授予单位】:天津大学
【学位级别】:硕士
【学位授予年份】:2012
【分类号】:TP391.41

【参考文献】

相关期刊论文 前2条

1 夏小玲;柴望;;基于Shape Context的形状匹配方法的改进[J];东华大学学报(自然科学版);2009年01期

2 郭斯羽;张煦芳;;一种基于模型的自适应阈值分割算法[J];浙江大学学报(工学版);2005年12期



本文编号:2073440

资料下载
论文发表

本文链接:https://www.wllwen.com/wenyilunwen/guanggaoshejilunwen/2073440.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户4fea0***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com