基于近似几何误差的动态隐式曲线重构

提出一种以代数张量积样条曲线作为几何表示形式,基于近似几何误差和薄板能量极小化的隐式重构模型.同时结合最优化理论中的信赖域思想，给出自适应的迭代求解算法及其实现．这种方法采取无代价初始化技术，通过迭代能稳定地达到目标点集的高质量重构，特别是对复杂形状的目标，具有很强的处理能力．     

基于隐式T样条的曲面重构算法

提出隐式T样条曲面，将T网格从二维推广到三维情形，同时利用八叉树及其细分过程，从无结构散乱点数据集构造T网格，利用曲面拟合模型将曲面重构问题转化为最优化问题；然后基于隐式T样条曲面将最优化问题通过矩阵形式表述，依据最优化原理将该问题转化成线性方程组，通过求解线性方程组解决曲面重构问题；最后结合计算实例进行讨论．该方法能较好地解决曲面重构问题，与传统张量B样条函数相比，能效地减少未知控制系数与计算量．     

有理B样条曲面的区间隐式化

提出有理B样条曲面的区间隐式化方法,即对一个有理B样条曲面,寻求包含给定的曲面的区间隐式B样条曲面,使得区间隐式B样条曲面的"厚度"尽量小,同时尽量避免出现多余分支.该问题等价于求区间隐式B样条曲面的2个边界曲面.针对该问题建立一个最优化模型并求解.     

层次隐式张量积B-样条曲面及其在曲面重构中的应用

提出一类新的隐式曲面表示形式,它具有良好的层次性与自适应性,特别适合于表示层次细节模型.首先给出层次隐式张量积B-样条曲面的定义,回顾曲面重构的一般数学模型,然后提出适于求解该类曲面的最优化模型以及定义域自适应分解的方法.在此基础之上,提出层次逼近算法来逼近采样点数据集.随后,引入单位分解方法,将其用于融合各个子域上的局部逼近函数,使其成为一个具有整体光滑性的函数.最后基于散乱点数据集,给出曲面重构的实例,并作简单的讨论.     

异度隐函数样条曲线曲面

隐式曲线曲面被广泛应用于曲线曲面插值、逼近与拼接. 通过添加辅助曲线曲面,提出异度隐函数样条曲线曲面方法,并对其插值性、凸性与正则性进行分析. 具体实例表明,异度隐函数样条提供了次数低、构造简单、灵活性好的曲线曲面插值与拼接方法.     

基于Kohonen神经网络的B样条曲面重构

探讨了三维散乱数据点的自由曲面自组织重构方法。建立了基于自组织特征映射神经网络的矩形网格重构模型及其训练算法。所建模型利用神经元对曲面散乱数据点的学习和训练来模拟曲面上点与点之间的内在关系,节点连接权向量集作为对散乱数据点集的工程近似化并重构曲面样本点的内在拓扑关系。通过该方法不仅能够对无规则散乱数据点进行逼近,并且通过该方法得到的曲面也可以作为后继曲面重构的初始曲面。仿真实验表明,所建神经网络模型可实现三维密集无规则数据点的曲面自组织重构集自压缩于一体。     

隐式磨光曲面的保凸性研究

对隐式代数磨光曲面的保凸性进行了研究。证明了如果两个隐式代数曲面是凸的,则它们光滑拼接后,得到的磨光曲面在一定条件下仍然是凸的。以凸三面角为例,提出了利用代数样条磨光的方法并证明了利用这种磨光方法得到的磨光曲面具有保凸性。     

基于B样条的平面轮廓重构闭合曲面算法

由一组平行轮廓线重构三维闭合表面是三维可视化研究的主要内容之一。文中通过对B样条插值算法的研究,提出了一种新的公共节点矢量确定方法,利用该方法首先对经过预处理的CT牙齿图片提取轮廓线获得三维数据点,之后对轮廓线数据点进行B样条曲线的拟合,在每条拟合曲线上根据所确定的节点矢量值重新采样,由重新采样的三维数据点利用B样条曲面插值算法构造闭合曲面．所构造的闭合曲面是对原始轮廓数据的拟合。通过实例验证可看出该方法可获得较好的拟合曲面,经过误差分析检测,满足拟合条件,因此该方法可以保证几何重建的准确性。     

一类三次隐式代数曲面参数化研究

在CAGD中隐式曲面和参数曲面作为曲面的两种表示形式各有其内在的优点 ,多年来如何有效地实现二者的相互转换一直是CAGD的一个热点问题 对一类GC1拼接两个二次曲面的三次混合代数曲面进行了研究 ,提出一种基于同轴平面束与代数曲面相交的几何化参数化方法 与传统参数化方法相比 ,该方法结构直观且具有可使三次代数曲面位于 [0 ,1]× [0 ,1]参数区间内 ,以及曲面的边界位于等参数线上等特点 ,利用这种参数曲面可以方便地实现机器作图和各种操作 实验结果验证了方法的有效性     

隐式T样条实现封闭曲面重建

为了简化法向偏差约束条件和优化光滑能量项,提出一种隐式T样条曲面重建算法.首先利用八叉树及其细分过程从采样点集构造三维T网格,以确定每个控制系数对应的混合函数;然后基于隐式T样条曲面建立目标函数,利用偏移曲面点集控制法向,采用广义交叉检验(GCV)方法估计最优光滑项系数,并依据最优化原理将该问题转化为线性方程组求解得到控制系数,从而实现三角网格曲面到光滑曲面的重建.在误差较大的区域插入控制系数进行T网格局部修正,使得重建曲面达到指定精度.该算法使重建曲面C1连续条件得到松弛,同时给出最优的光顺项系数估计,较好地解决了封闭曲面的重建问题.实例结果表明,文中算法逼近精度高,运算速度快,仿真结果逼真.     

Real-time ray casting of algebraic B-spline surfaces

Piecewise algebraic B-spline surfaces (ABS surfaces) are capable of modeling globally smooth shapes of arbitrary topology. These can be potentially applied in geometric modeling, scientific visualization, computer animation and mathematical illustration. However, real-time ray casting the surface is still an obstacle for interactive applications, due to the large amount of numerical root findings of nonlinear polynomial systems that are required. In this paper, we present a GPU-based real-time ray casting method for ABS surfaces. To explore the powerful parallel computing capacity of contemporary GPUs, we adopt iterative numerical root-finding algorithms, e.g., the Newton-Raphson and regula falsi algorithms, rather than recursive ones. To facilitate convergence of the Newton-Raphson or regula falsi algorithm, their initial guesses are determined through rasterization of the isotopic isosurface, and the isosurface is generated based on regular criteria for surface domain subdivision. Meanwhile, polar surfaces are adopted to identify single roots or to isolate different roots, i.e., ray and surface intersections. As an important geometric feature, the silhouette curve is elaborately computed to floating-point accuracy, which can be applied in further anti-aliasing processes. The experimental results show that the proposed method can render thousands of piecewise algebraic surface patches of degrees 6-9 in real time.     

基于轮廓数据的B样条曲面重建

针对B样条曲面拟合中出现的问题和困难,提出了一种基于行组织的轮廓数据（截面数据）的曲面重建方法。该方法避免了数据点的参数化问题,使得逼近曲面拥有较好的形状和合理的控制顶点数量。该方法的基本思想是：首先构造易于控制的低阶曲面拟合数据点,此曲面称控制曲面,然后利用高次曲面逼近该曲面,此高次曲面称为逼近曲面,为所需要的重建曲面。在曲面重建中利用最佳平方逼近和光顺函数,减少了逼近曲面的控制顶点冗余,较有效地防止了逼近曲面的形状突变和曲面的扭曲,很大程度地提高了曲面的质量。     

基于力密度的交互式B样条曲面变形

由于传统曲面造型方法缺乏直观调控及难于交互的局限性,强调了基于力学平衡和力密度方法的物理造型策略,通过建立外载荷变化和节点位移等物理意义的最小化线性方程从而实现B样条曲面的自适应性变形调整。系统地给出了基于各种几何约束下通用的几何求解方法,并进一步提出了交互式目标曲线与区域曲线自由绘制的曲面整体和局部变形技术。实验结果验证了此方法的高效性。     

Curve reconstruction based on an interval B-spline curve

Curve reconstruction that generates a piece of centric curve from a piece of planar strip-shaped point cloud is a fundamental problem in reverse engineering. In this paper, we present a new curve-reconstruction algorithm based on an interval B-spline curve. The algorithm constructs a rectangle sequence approximating the point cloud using a new data clustering technique, which facilitates the determination of curve order implied in the shape of the point cloud. A quasicentric point sequence and two pieces of boundary point sequences are then computed, based on which a piece of interval B-spline curve representing the geometric shape of the point cloud is constructed. Its centric curve is the final reconstructed curve. The whole algorithm is intuitive, simple, and efficient, as demonstrated by experimental results.     

任意散乱点集的B-样条曲面重建

为了解决工业设计中复杂形体的曲面造型问题,提出了一种张量积型的低阶B-样条曲面重建算法。先将采集到的任意拓扑形状的散乱数据点进行三次不同的参数化得到四边形控制网格,然后再采用张量积型的双二次、双三次B-样条进行拟合,在拟合的过程中采用距离函数来控制拟合误差,得到光滑的曲面。运用该方法,直接对初始散乱点集进行重建,方法简单易实施,重建效率高并且重建后的样条曲面自然满足切平面连续。与以往的方法相比,该方法在逆向工程中可以在保证连续性的情况下,得到精准的结果曲面,提高了曲面造型的质量和效率。     

B-spline surface fitting based on adaptive knot placement using dominant columns

By expanding the idea of B-spline curve fitting using dominant points (Park and Lee 2007 [13]), we propose a new approach to B-spline surface fitting to rectangular grid points, which is based on adaptive knot placement using dominant columns along u- and v-directions. The approach basically takes approximate B-spline surface lofting which performs adaptive multiple B-spline curve fitting along and across rows of the grid points to construct a resulting B-spline surface. In multiple B-spline curve fitting, rows of points are fitted by compatible B-spline curves with a common knot vector whose knots are computed by averaging the parameter values of dominant columns selected from the points. We address how to select dominant columns which play a key role in realizing adaptive knot placement and thereby yielding better surface fitting. Some examples demonstrate the usefulness and quality of the proposed approach.     

基于广义多项式神经网络的点云数据隐式曲面重构方法

针对点云数据的三维重建问题,提出了一种隐曲面重构的广义多项式神经网络新方法.该广义多项式神经网络隐层各神经元激励函数互不相同且线性无关,能够对应地学习点云数据样本中不同的模式,因此,具有较好的学习能力.基于梯度下降法原理,推导了其学习算法.仿真实验尝试将该方法应用于一些简单封闭物体的带噪点云数据隐式曲面重建,取得了较理想的重建质量和去噪效果.     

Functional networks for B-spline surface reconstruction

Recently, a new extension of the standard neural networks, the so-called functional networks, has been described [E. Castillo, Functional networks, Neural Process. Lett. 7 (1998) 151–159]. This approach has been successfully applied to the reconstruction of a surface from a given set of 3D data points assumed to lie on unknown Bézier [A. Iglesias, A. Gálvez, Applying functional networks to CAGD: the tensor-product surface problem, in: D. Plemenos (Ed.), Proceedings of the International Conference on Computer Graphics and Artificial Intelligence, 3IA’2000, 2000, pp. 105–115; A. Iglesias, A. Gálvez, A new artificial intelligence paradigm for computer-aided geometric design, in: Artificial Intelligence and Symbolic Computation, J.A. Campbell, E. Roanes-Lozano (Eds.), Lectures Notes in Artificial Intelligence, Berlin, Heidelberg, Springer-Verlag, vol. 1930, 2001, pp. 200–213] and B-spline tensor-product surfaces [A. Iglesias, A. Gálvez, Applying functional networks to fit data points from B-spline surfaces, in: H.H.S. Ip, N. Magnenat-Thalmann, R.W.H. Lau, T.S. Chua (Eds.), Proceedings of the Computer Graphics International, CGI’2001, IEEE Computer Society Press, Los Alamitos, CA, 2001, pp. 329–332]. In both cases the sets of data were fitted using Bézier surfaces. However, in general, the Bézier scheme is no longer used for practical applications. In this paper, the use of B-spline surfaces (by far the most common family of surfaces in surface modeling and industry) for the surface reconstruction problem is proposed instead. The performance of this method is discussed by means of several illustrative examples. A careful analysis of the errors makes it possible to determine the number of B-spline surface fitting control points that best fit the data points. This analysis also includes the use of two sets of data (the training and the testing data) to check for overfitting, which does not occur here.     

Reconstruction of B-spline surface by interpolating boundary curves and approximating inner points

In this paper,we present an algorithm for reconstruction of B-spline surface such that it interpolates the four given boundary curves and simultaneously approximates some given inner points.The main idea of our method is:first,we construct an initial surface which interpolates the four given boundary curves;then,while keeping the boundary control points of the initial surface unchanged,we reposition the inner control points of the surface with energy optimization method.Examples show that our algorithm is practicable and effective.