四元数在关键帧骨骼动画中的应用

在关键帧骨骼动画中,中间帧的生成效果直接影响着动画的自然流畅程度。生成中间帧的方法有线性插补法和高次多项式拟合法等。由于人体的骨骼关节之间存在着连带关系,且关节的旋转角度受到约束,对人体的关键帧姿态进行插值不同于一般物体。本文在建立人体骨骼模型的基础上,采用四元数方法对给定的关键帧进行插值生成中间帧,并对生成结果做简单的分析。     

关键帧动画物体变形动画和过程动画

一、关键帧动画关键帧的概念来源于传统的卡通片制作。在早期WaltDisney的制作室，熟练的动画师设计卡通片中的关键画面，也即所谓的关键帧，然后由一般的动画师设计中间帧。在三维计算机动画中，中间顿的生成由计算机来完成，插值代替了设计中间帧的动画师。所有影响画面图像的参数都可成为关键帧的参数，如位置、旋转角、纹理的参数等。关键顿技术是计算机动画中最基本并且运用最广泛的方法。另外一种动画设置方法是样条驱动动画。在这种方法中，用户指定物体运动的轨迹样条，轨迹样条由用户交互给出。几乎所有的动画软件如Alias、bofti…     

基于关键帧的三维人脸皱纹动画

表现人脸的皱纹特征是提高人脸动画真实感的重要因素之一,文章提出了一种基于关键帧的皱纹动画方法,使用高度图、法线图和MPEG-4人脸运动参数描述皱纹动画关键帧,通过对高度图、法线图插值生成皱纹动画中间帧。所提方法对人脸模型网格复杂度要求不高,且合成的皱纹动画具有真实感强和实时性高的特点。     

一种视频驱动的三维虚拟人动画实现方法

研究实现三维人体动画具有广泛的应用前景和实用意义,提出了一种二维视频驱动的三维人体动画实现方法。基于动态帧的关键帧提取算法从二维视频中构建了二维关键帧集合;基于二维关键帧构建二维人体骨骼模型;利用小孔成像原理和勾股定理计算得到关节特征点的深度坐标,从而得到了反映人体动画的三维数据。实验结果表明,该方法生成的三维人体动画形象逼真、成本低、提高了运动生成的实时性,能够应用于虚拟现实、计算机游戏、三维视频游戏制作等领域。     

计算机二维动画的算法实现

本文基于关键帧控制的动画原理，提出了一种基于多形描述目标，并通过多边形对主顶点插补中间帧的计算机二维动画算法。     

手绘艺术动画的计算机辅助设计方法综述

近年来随着计算机技术广泛应用于手绘艺术动画的制作,出现了许多新的手绘艺术动画的计算机辅助设计和制作方法。这些方法概括为五类：①结合三维动画技术实现艺术动画制作;②引入中国传统的艺术元素制作手绘动画;⑧基于视频流生成动画;,④使用素材重用的方法制作手绘动画;⑤结合中间帧自动或半自动生成技术制作动画。文章对这些方法进行了介绍,并指出中间帧生成技术能够保持绘画者的手绘风格,突出手绘动画的艺术特征;并进一步说明了结合中间帧生成技术辅助制作动画的关键,是寻求一种自然高效的交互方式来定义始末关键帧中图形元素间的对应关系,以及根据此对应关系自动或半自动生成中间帧的方法。     

一种基于分形的关键帧算法研究

提出了一种基于分形的关键帧算法,算法对初始关键帧和终止关键帧之间的压缩映射的对应准则和概率的对应准则各给出了一种较好的解决方法.实验结果表明,算法产生的动画过渡平滑,节奏自然,画面流畅,本文算法的主要特征是它能处理传统的关键帧算法所不能处理的分形物体,在实际应用中,该算法能够在提高制作效率和中间帧质量的同时大大减少动画制作的成本.     

圆弧路径法，一种新的多边形变形方法

二维形状变形技术在二维角色动画、模式匹配、几何造型、虚拟现实、工业模拟、科学计算可视化等领域有着重要的应用。本文提出了一种顶点路径圆弧法的二维形状变形新方法。该算法通过控制关键帧多边形顶点按照一条特殊的圆弧曲线路径进行运动，实现二维形状变形。通过许多实例的测试表明，该算法效果良好：不仅可以保持首末关键帧形状的共同特征，而且中间插值形状变化自然平滑。同时，我们的方法易于用户交互控制；容易推广到高维情形；计算量较小、能达到系统实时的要求。     

计算机动画关键帧插补技术综述

对计算机辅助动画制作过程中关键帧的插补技术进行了研究。线性插值法通过给定的起始帧和结束帧对应点之间的直线距离算出中间插值帧。移动点约束条件插补法利用随时间空间变化的曲线与彩体上的点相关联，来控制关键帧上对应控制点的轨迹以厦动态特征。骨架插补法是将图形抽象成骨架，然后进行插补。     

基于线型草图的多源3维动画传输

为了对3维动画进行鲁棒地传输,提出了一种基于线型草图的3维动画传输方法,该方法可以把多个源网格的动画传输到一个目标网格。该方法由以下几个步骤组成:(1)使用线型草图在源网格和目标网格间建立映射;(2)使用差分均值坐标(differential mean value coordinates),根据源动画来变形目标网格;(3)对齐各部分变形后的目标网格,通过求解最小化约束函数,以获取光滑的目标网格,从而生成目标动画的关键帧;(4)对由步骤(3)产生的结果进行插值,即可产生完整的目标对象的3维动画。通过对上述几个方面的深入研究,以及将这些成果集成在原型系统中的结果表明,该方法具有以下特点:①既不需要源网格和目标网格有相同的顶点数和三角面片数,也不需要有类似的拓扑信息;②源动画既可以是3维的,也可以是2维视频或Flash;③目标对象可以是无结构化的点云数据。实践表明,该方法直观易用,可以生成逼真的3维动画。最后通过从狗、猫、蛇、人体的3维动画与2维车轮视频的合成来产生复杂的马车动画,对该方法进行了验证。另外,原型系统还通过产生其他几个例子来进一步验证其通用性和鲁棒性。     

Segmenting a deforming mesh into near-rigid components

Given a deforming mesh in an animation, we propose a new method to segment this mesh into several near-rigid sub-meshes. From this deforming mesh over all frames of an animation, we can analyze the degree of deformation between two nearby faces on the mesh. Then, our algorithm partitions the given deforming mesh into near-rigid components where the segmentation boundaries always pass at regions of large deformation. As a result, the mesh segmentation is invariant to all frames of the given animation and the motion of faces in each near-rigid-component can be represented by the same approximate affine transformation. To demonstrate the usefulness of the algorithm, we solve the restriction of deformation transfer for triangle meshes [31] which requires similar reference poses between source mesh and target mesh.     

Feature matching and affine transformation for 2D cell animation

This paper presents an application for feature matching and affine transformation in computer graphics. The study case considered is 2D cell animation, which is still a labor-intensive process in current cartoon film production. One key problem in automating 2D cell animation in inbetween frame generation. The objective of our work is to investigate how to automatically generate inbetweening from pairs of hand-drawn 2D key frames. Our technique first establishes the correspondence of feature points and then estimates the affine transformation between each pair of 2D key frames. The inbetween frames are then automatically generated by interpolation.     

基于QPSO的多模态医学图像配准研究

本文提出一种基于QPSO的形态金字塔多模态医学图像配准算法,该方法以梯度互信息作为相似性度量标准,运用全局仿射变换建立数学模型,充分利用形态金字塔的非线性、多分辨率等特性将图像进行分层处理,并且利用QPSO和Powell算法结合逐层获取参数解。实验结果表明在任意初始值的情况下,基于QPSO的寻优算法比Levenberg-Marquardt寻优算法运行速度快,并且配准精度更高。     

A Colour Interpolation Scheme for Topologically Unrestricted Gradient Meshes

Gradient meshes are a 2D vector graphics primitive where colour is interpolated between mesh vertices. The current implementations of gradient meshes are restricted to rectangular mesh topology. Our new interpolation method relaxes this restriction by supporting arbitrary manifold topology of the input gradient mesh. Our method is based on the Catmull‐Clark subdivision scheme, which is well‐known to support arbitrary mesh topology in 3D. We adapt this scheme to support gradient mesh colour interpolation, adding extensions to handle interpolation of colours of the control points, interpolation only inside the given colour space and emulation of gradient constraints seen in related closed‐form solutions. These extensions make subdivision a viable option for interpolating arbitrary‐topology gradient meshes for 2D vector graphics.     

The Direction of Synthetic Actors in the Film Rendez-Vous a Montreal

This article describes how the synthetic actors Marilyn Monroe and Humphrey Bogart were directed in the film Rendez-vous a Montreal, using the Human Factory software. Innovative points are emphasized. The recreation of famous persons and reconstitution of their personalities, as well as the corresponding animation, are considered from an artistic point of view. Technical points discussed are personalized expressions, abstract muscle action (AMA) procedures, expression key-frame interpolation, multitrack mixing and synchronization, automatic body mapping using joint-dependent local deformation (JLD) operators, and semi-automatic object grasping. Task-level problems are also explained. Finally, how a fifth-generation animation system might evolve is discussed.     

基于单幅图像的真实感人体动画合成

由于单幅图像缺失三维信息以及完整的纹理信息,基于单幅图像的真实感三维人体动画合成极具挑战性。针对单幅图像三维信息缺失问题,提出了一种基于SMPL参数模型的三维人体几何重建方法。该方法以单幅图像为输入,先根据输入图像人体轮廓信息变形标准的SMPL参数模型分别生成与目标轮廓一致的正反面的三维几何模型,然后利用基于B样条插值的网格拼接融合算法拼接正反面三维几何,最后为了恢复正确的手部几何,利用基于B样条插值的网格拼接融合算法,将重建后的模型上错误的手部几何用标准SMPL参数模型上正确的手部几何替换。同时,针对单幅图像中纹理缺失的问题,提出了一个称为FBN（front to back network）的对抗生成网络,用于恢复被遮挡的人体背面纹理。实验结果表明,该方法生成的具有完整纹理的人体几何能够由3D运动数据驱动运动,生成具有高度真实感的三维人体动画。     

邻域表示的形状刚性操作实现方法

角色动画编辑的核心技术之一,是形状既能实现刚性变形,又能实现实时操作。在形状刚性操作计算中,通过定义三角网格顶点的面邻域和点邻域的平方差度量,简化自由顶点与约束顶点的坐标分离。这种表示法,在整体拟合中使得x坐标与y坐标能够独立求解。在算法实现中,设计恰当的系数矩阵和常数项向量的链表结构,采用矩阵索引存储和共轭梯度求解,减少数据的重复计算。实验表明,在普通的PC机上能够实现约1 000个顶点的网格实时交互变形操作。     

Mesh deformation using radial basis functions for gradient-based aerodynamic shape optimization

Gradient-based aerodynamic shape optimization using computational fluid dynamics (CFD), and time dependent problems in aeroelasticity, that is, coupled calculations between computational structural mechanics (CSM) and CFD, require repeated deformations of the CFD mesh.An interpolation scheme, based on radial basis functions (RBF), is devised in order to propagate the deformations from the boundaries to the interior of the CFD mesh. This method can lower the computational costs due to the deformation of the mesh, in comparison with the usual Laplace smoothing. Moreover, the algorithm is independent of the mesh connectivities. Therefore, structured and unstructured meshes are equally treated as well as hybrid meshes.The application of this interpolation scheme in problems of aerodynamic shape optimization is also carefully investigated. When the optimization is executed by a gradient-based algorithm the cost function is differentiated with respect to the design parameters in order to obtain the gradient. The gradient is most efficiently and accurately calculated by solving a certain adjoint equation derived from the discretized flow equations. The calculation of the gradient, which is detailed in this presentation, involves the Jacobian matrix of the mesh deformation.Finally, we present the results of an optimization of the ONERA M6 wing at transonic speed using the interpolation algorithm. The results are used for comparison with another technique of mesh deformation. The quality of the mesh obtained by the new algorithm, and the interpolation error, are analyzed with respect to the parameters of the interpolation scheme: the type of RBF, the RBF’s shape parameter, and the sets of control points.     

Mesh Sequence Morphing

Morphing is an important technique for the generation of special effects in computer animation. However, an analogous technique has not yet been applied to the increasingly prevalent animation representation, i.e. 3D mesh sequences. In this paper, a technique for morphing between two mesh sequences is proposed to simultaneously blend motions and interpolate shapes. Based on all possible combinations of the motions and geometries, a universal framework is proposed to recreate various plausible mesh sequences. To enable a universal framework, we design a skeleton‐driven cage‐based deformation transfer scheme which can account for motion blending and geometry interpolation. To establish one‐to‐one correspondence for interpolating between two mesh sequences, a hybrid cross‐parameterization scheme that fully utilizes the skeleton‐driven cage control structure and adapts user‐specified joint‐like markers, is introduced. The experimental results demonstrate that the framework, not only accomplishes mesh sequence morphing, but also is suitable for a wide range of applications such as deformation transfer, motion blending or transition and dynamic shape interpolation.     

一种基于多层解析的MPEG-4二维网格动画水印技术

提出了一种针对MPEG-4二维网格动画的强健数字水印系统的实现方法：首先利用多层解析分析将二维网格动画中的运动特征提取出来，再以扩频技术将水印信息分散到这些运动特征中，同时利用空间平面校正技术来解决几何转换所造成的网格位置偏移问题。使用了仿射变换攻击、平滑化攻击、强化及衰减攻击、随机噪声攻击以及混合性攻击来测试生成的水印系统。实验结果显示该水印系统实现方法能有效地抵抗上述攻击。