逼近散乱点云数据的三角形网格精确剖分

基于自组织特征映射神经网络构建的三角形网格模型可以实现测量点云 压缩后的Delaunay 三角逼近剖分,但该模型存在逼近误差和边缘误差。为减小三角形网格 的逼近误差和边缘误差,构建了精确逼近的三角形网格模型。首先采用整个测量点云,对三 角形网格模型中的所有神经元进行整体训练;然后对三角形网格中的网格神经元的位置权 重,沿网格顶点法矢方向进行修正;最后采用测量点云中的边界点集,对三角形网格模型中 的网格边界神经元进行训练。算例表明,应用该模型,可以有效减小三角形网格的边缘误差, 三角形网格逼近散乱点云的逼近精度得到大幅提高并覆盖散乱点云整体分布范围。     

基于SOM的散乱数据点集的B样条曲面重建

利用自组织映射神经网络(SOM)技术对散乱数据点集进行B样条曲面重建时,往往存在网络学习时间过长和学习效果不理想等问题。提出了一种新的神经元初始化方法和分块学习算法,该算法首先运用主元素分析方法(PCA)对散乱数据进行分块,将拓扑结构为四边形的输出层神经元初始化在每块散乱数据的最小二乘平面上进行网络学习和训练,将分块学习得到的各网格曲面拼接成一个整体;然后对该整体网格曲面的边界和内部单独学习,得到一张逼近待重建曲面的双线性B样条曲面;最后对该B样条曲面误差进行了修正。实例证明,该算法可以明显地减少SOM网络学习时间,并改善网络学习效果。     

一种基于扫描点的平面散乱点云边界的提取算法

根据平面散乱点云的特点,依据边界点和非边界点的位置关系,用网格把点云划分开来,在确定边界网格和非边界网格以后,将所有的边界网格按照它们的位置关系连接成环,对于从每一个边界网格提取边界提出了一种最小凸边算法,并且从理论和实验上证明了这种最小凸边算法的可行性。     

散乱点云的三角划分算法研究

通过对当前的三角网格划分方法进行比较分析,提出了一种散乱点云的3D三角网格划分算法。该算法不需如同二维划分方法那样要对散乱点云对应的自由曲面分片投影,而是直接在3D空间,根据离散点集所对应的曲面形态变化,利用网格扩展、边界环分裂和边界环封闭,逐层收缩生成三角网格。该算法能方便地处理空间多种曲面的散乱点云数据,并且生成的三角网格形态优良,布局合理。     

保留边界的点云简化方法

针对点云简化算法中边界点丢失的问题,提出了一种保留边界的三维散乱点云的非均匀简化算法。首先利用kd-tree建立散乱数据点云的空间拓扑关系,计算出每个数据点的k邻域;然后针对目前依据点云分布均匀性算法提取边界效率低的问题,提出一种改进的点云边界点判定算法;最后保留所有边界点,对非边界点,根据曲面变分值和k邻域点已保留比例,进行点云的非均匀简化。实验结果表明,该算法精度高,空间复杂度低,而且简化后点云边界保留完整。     

基于Hermite插值的网格拼接和融合

网格模型的拼接和融合是3维形状编辑和造型中的一个重要方面。基于Hermite插值技术,提出一种适用于具有一般边界点空间分布的三角网格模型之间无缝光滑拼接和融合方法。首先查找网格模型待拼接区域的边缘点集,并利用二次B样条曲线插值边缘点集分别得到边缘曲线;然后对边缘曲线进行Hermite插值得到拼接区域连续曲面;最后对拼接曲面分别进行三角网格化和Laplacian光顺平滑处理以实现网格模型的光滑拼接和融合。由于利用B样条曲线插值待拼接模型边界,本文方法适用于具有各种不同边界情形的网格模型拼接和融合,它不仅仅可以处理平面边界曲线情形也可以处理空间边界曲线情形。结合Hermite曲面插值拼接过渡区域,使得产生的拼接网格能光滑地衔接待拼接模型。实验结果表明,本文方法能够有效地实现三角网格模型的光滑拼接、模型修复和模型融合。     

基于Hermite插值的网格拼接和融合-CIDE2013

网格模型的拼接和融合是3维形状编辑和造型中的一个重要方面。基于Hermite插值技术,提出了一种适用于具有一般边界点空间分布的三角网格模型之间无缝光滑拼接和融合方法。首先查找网格模型待拼接区域的边缘点集,并利用二次B样条曲线插值边缘点集分别得到边缘曲线;然后对边缘曲线进行Hermite插值得到拼接区域连续曲面;最后对拼接曲面分别进行三角网格化和Laplacian光顺平滑处理以实现网格模型的光滑拼接和融合。由于利用B样条曲线插值待拼接模型边界,本文方法适用于具有各种不同边界情形的网格模型拼接和融合,它不仅仅可以处理平面边界曲线情形也可以处理空间边界曲线情形。结合Hermite曲面插值拼接过渡区域,使得产生的拼接网格能光滑地衔接待拼接模型。实验结果表明,本文方法能够有效地实现三角网格模型的光滑拼接、模型修复和模型融合。     

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。     

心内膜散乱点云边界点检测算法研究

针对心内膜散乱点云预处理中的边界点检测,利用截线云理论将散乱点云进行等间隔区域分层,将点云投影至点云切片,得到切片的散乱点集,同时建立链表结构分区存储点云数据;由平面上点的二维坐标定位,提出区域"十"字算法进行切片数据边界点提取,获取切片数据的最外层点,将检测到的边界点存回原始三维数据源,完成预处理过程。实验结果证明,该算法对边界点具有较强的识别能力,能够在快速、有效地简化点云数据的同时保持原始特征的信息,可以提高后续三维建模的精度和速度。     

散乱点云的快速增量网格重建算法

散乱数据的网格重建是数字几何处理的基础性技术之一.本文提出一种快速增量式散乱点云网格重建算法,运用波前( Wave Front)方法渐进地由点云数据生成物体表面的网格模型.该算法以一个”种子”三角形初始化搜索队列,以逐渐生成的新边为搜索元素,借助Kd-树空间划分技术和搜索约束条件,快速完成优化点的评估及三角面片重建,可在保证网格质量的同时,过滤部分对重建效果意义不大的点.实验表明,该算法能够高效、可靠地生成具有不同几何复杂度的原始曲面二维流形三角网格逼近,适用于海量数据点的网格重建.     

Generating a smooth voxel-based model from an irregular polygon mesh

A method for generating a smooth voxelbased model from an arbitrary polygon mesh is presented. It is based on a polygonal subdivision process which takes an irregular polygon mesh as input and creates a finer and smoother mesh. The mesh is recursively refined down to or close to the voxel level, and then voxelized (digitized) into a voxel-based representation. A local subdivision approach has been developed in order to ease the computationally expensive subdivision process. The voxelization of the mesh maintains topological and fidelity requirements which are pre-defined and application dependent.     

Classification-Based Probabilistic Modeling of Texture Transition for Fast Line Search Tracking and Delineation

We introduce a classification-based approach to finding occluding texture boundaries. The classifier is composed of a set of weak learners which operate on image intensity discriminative features which are defined on small patches and fast to compute. A database which is designed to simulate digitized occluding contours of textured objects in natural images is used to train the weak learners. The trained classifier score is then used to obtain a probabilistic model for the presence of texture transitions which can readily be used for line search texture boundary detection in the direction normal to an initial boundary estimate. This method is fast and therefore suitable for real-time and interactive applications. It works as a robust estimator which requires a ribbon like search region and can handle complex texture structures without requiring a large number of observations. We demonstrate results both in the context of interactive 2-D delineation and fast 3-D tracking and compare its performance with other existing methods for line search boundary detection.     

基于三角网格构建的扫描测头半径三维补偿

采用扩展的自组织特征映射神经网络探讨了三坐标测量机接触式密集数据采集的测头半径三维补偿。构建了基于三角形网格构建的测头半径三维补偿模型。经过训练,神经网络将整个数字化点群数据分成许多子区域,每个子区域用一个微切平面逼近;对子区域的分类核心,即神经元位置权重,沿微切平面法矢方向进行修正,得到逼近测头球心面的三角形网格II;根据微切平面的法线,对测头半径进行三维补偿,得到逼近接触曲面的三角形网格III。测头半径三维补偿的法矢方向,也可通过估算三角网格II顶点的法矢得到。算例表明所创建的测头半径三维补偿模型有效可行。     

Reconstructing regular meshes from points

We propose an algorithm for reconstructing regular meshes from unorganized point clouds. At first, a nearly isometric point parameterization is computed using only the location of the points. A mesh, composed of nearly equilateral triangles, is later created using a regular sampling pattern. This approach produces meshes with high visual quality and suitable for use with applications such as finite element analysis, which tend to impose strong constraints on the regularity of the input mesh. Geometric properties, such as local connectivity and surface features, are identified directly from the points and are stored independent of the resulting mesh. This decoupling preserves most details and allows more flexibility for meshing. The resulting parameterization supports several direct applications, such as texturing and bump mapping. In addition, novel boundary identification and cut parameterization algorithms are proposed to overcome the difficulties caused by cuts, non-closed surfaces and possible self-overlapping parameter patches. We demonstrate the effectiveness of our approach by reconstructing regular meshes from real datasets, such as a human colon obtained from CT scan and objects digitized using laser scanners.     

飞行时间3维相机的多视角散乱点云优化配准

针对目前基于飞行时间(TOF)原理的3维相机实现物体完整表面的3维点云重建过程中,多视角散乱点云配准精度低的问题,提出一种优化配准方法。该方法通过构建一个目标功能函数,并结合相邻点云的转换矩阵对该目标函数进行最小化求解,直接获取任意位置的点云到基准点云所处坐标系的绝对转换矩阵,避免了对连续点云之间的配准而引起误差的累加。对不同的物体进行实验,实验结果表明,该方法在保证点云配准速度的同时,提高了多视角点云配准的精度,物体点云模型重建效果较好,有利于实现后期3维曲面网格的重建。     

Automatic hexahedral mesh generation for multi-domain composite models using a hybrid projective grid-based method

This paper presents an algorithm to generate an all-hexahedral mesh of a multi-domain solid model using a hybrid grid-based approach. This is based on a projective concept during the boundary adaptation of the initial mesh. In general, the algorithm involves the generation of a grid structure, which is superimposed on the solid model. This grid structure forms an initial mesh consisting of hexahedral elements, which intersect fully or partially with the solid model. This initial mesh is then shrunk in an outside-in manner to the faces of the model through a node projection process using the closest position approach. To match the resulting mesh to the edges of the model, a minimal deformation angle method is used. Finally, to match the vertices with the nodes on the mesh, a minimal warp angle method is employed. To create the mesh of a multi-domain solid model, an outside-in and inside-in hybrid of the grid-based method is used. This hybrid method ensures that the meshes of the different domains are conforming at their common boundary. This paper also describes two methods for resolving cases of degenerate elements: a splitting technique and a wedge insertion technique.     

Direct boolean intersection between acquired and designed geometry

In this paper, a new shape modeling approach that can enable direct Boolean intersection between acquired and designed geometry without model conversion is presented. At its core is a new method that enables direct intersection and Boolean operations between designed geometry (objects bounded by NURBS and polygonal surfaces) and scanned geometry (objects represented by point cloud data).We use the moving least-squares (MLS) surface as the underlying surface representation for acquired point-sampled geometry. Based on the MLS surface definition, we derive closed formula for computing curvature of planar curves on the MLS surface. A set of intersection algorithms including line and MLS surface intersection, curvature-adaptive plane and MLS surface intersection, and polygonal mesh and MLS surface intersection are successively developed. Further, an algorithm for NURBS and MLS surface intersection is then developed. It first adaptively subdivides NURBS surfaces into polygonal mesh, and then intersects the mesh with the MLS surface. The intersection points are mapped to the NURBS surface through the Gauss-Newton method.Based on the above algorithms, a prototype system has been implemented. Through various examples from the system, we demonstrate that direct Boolean intersection between designed geometry and acquired geometry offers a useful and effective means for the shape modeling applications where point-cloud data is involved.     

Three-Dimensional Measurement Using Structured Light Based on Deep Learning

Three-dimensional (3D) reconstruction using structured light projection has the characteristics of non-contact, high precision, easy operation, and strong real-time performance. However, for actual measurement, projection modulated images are disturbed by electronic noise or other interference, which reduces the precision of the measurement system. To solve this problem, a 3D measurement algorithm of structured light based on deep learning is proposed. The end-to-end multi-convolution neural network model is designed to separately extract the coarse- and fine-layer features of a 3D image. The point-cloud model is obtained by nonlinear regression. The weighting coefficient loss function is introduced to the multi-convolution neural network, and the point-cloud data are continuously optimized to obtain the 3D reconstruction model. To verify the effectiveness of the method, image datasets of different 3D gypsum models were collected, trained, and tested using the above method. Experimental results show that the algorithm effectively eliminates external light environmental interference, avoids the influence of object shape, and achieves higher stability and precision. The proposed method is proved to be effective for regular objects.