A two-step outlier removal procedure for surface reconstruction

Outliers in point clouds affect the performance of surface reconstruction directly. Most of outlier removal methods just remove those outliers far away from the real surface and are only applied to handle watertight surface. In this paper, a two-step outlier removal procedure is proposed to filter the point clouds acquired from the gray code and line-shifting technique. The first step is to remove the outliers far away from the real surface. Some feature points are extracted from the point clouds to construct an initial surface. The points with distances to the initial surface greater than a given threshold are removed as distant outliers. The retained points are linked into lines in each structured light sheet using their Voronoi diagrams. Some of lines which are very close to the real surface are removed as near outliers in the second step. The experimental results show that the proposed method is very effective in removing outliers for surface reconstruction. Foundation item: the National Natural Science Foundation of China (No. 30470488)     

基于结构光的多投影显示系统图像对准算法

为了解决多投影显示系统的图像对准问题,提出了一种基于结构光投影技术的对准算法.该算法通过求取等相位线的交点来建立投影机图像与相机拍摄图像间的映射关系,对于投影机图像上任一像素点,有水平和垂直2个相位值.在相机图像的水平相位展开图中找出与该点水平相位值相等的离散坐标点的位置,拟合出一条水平等相位线,同理,可以得到它的一条垂直等相位线;求出水平和垂直等相位线的交点,即为该点对应于相机图像中的像素点.该算法无需知道显示墙的解析表达式和相机的内部参数,适用于任意光滑的曲面显示墙.对准后图像与原始图像的对比实验表明:峰值信噪比达到27.904,结构相似度为0.972 4,说明该算法能有效提高图像的对准精度.      

Fast triangulation and local optimization for scan data of laser radar

In order to study the triangulation for the point cloud data collected by three-dimensional laser radar, in accordance with the line-by-line characteristics of laser radar scanning, an improved Delaunay triangulation method is proposed to mesh the point cloud data as a triangulation irregular network. Based on the geometric topology location information among radar point cloud data, focusing on the position relationship between adjacent scanning line of the point data, a preliminary match network is obtained according to their geometric relationship. A reasonable triangulation network for the object surface is acquired after the use of local optimization on initial mesh by Delaunay rule. Meanwhile, a new judging rule is proposed to contrast the triangulation before and after the optimization on the network. The result shows that triangulation for point cloud with full use of its own characteristics can improve the speed of the algorithm obviously, and the rule for judging the triangulation can evaluate the quality of network.     

Calibration for 3D Structured Light Measurement

A calibration procedure was developed for three-dimensional(3D) binocular structured light measurement systems. In virtue of a specially designed pattern, matching points in stereo images are extracted. And then sufficient 3D space points are obtained through pairs of images with the intrinsic and extrinsic parameters of each camera estimated prior and consequently some lights are calibrated by means of multi point fitting. Finally, a mathematical model is applied to interpolate and approximate all dynamic scanning lights based on geometry. The process of calibration method is successfully used in the binocular 3D measurement system based on structured lights and the 3D reconstruction results are satisfying.     

Novel mesh technique and its application in the wind field simulation for flexible spatial structure

In this paper, novel mesh techniques are proposed for wind field simulation of flexible spatial structure. For mesh generation, an interpolation strategy is presented to obtain a mesh system with variable density. Two spatial structure examples are used to examine the efficiency and applicability of this technique. Then based on the structured mesh system generated by the technique, the mesh nodal coordinates are updated to adapt the moving boundary conditions by means of the mapping interpolation functions and some examples are given to verify the effectiveness. Furthermore, the constrained counterforce distribution technique and projection interpolation strategy are developed to implement the data exchange on the interaction surface of wind and structure. Finally, the computational accuracy is numerically validated.     

Mesh motion approach based on spring analogy method for unstructured meshes

Mesh motion strategy is one of the key points in many fluid-structure interaction problems. One popular technique used to solve this problem is known as the spring analogy method. In this paper a new mesh update approach based on the spring analogy method is presented for the effective treatment of mesh moving boundary problems. The proposed mesh update technique is developed to avoid the generation of squashed invalid elements and maintain mesh quality by considering each element shape and grid scale to th...     

基于一面两孔特征的点云配准方法

为了提高零件在扫描检测过程中点云与设计模型的配准精度,提出了一种基于一面两孔特征的点云配准方法.该方法粗配准以零件的平面/圆柱孔特征为对象,使设计模型和点云的局部坐标系重合,并通过改进ICP算法求解点云与设计模型最近点的距离最小平方和实现精配准.由于配准区域和最近点的计算方法不同,精配准进一步分为全域和特征域配准两种类型.全域精配准以距点云最近的设计模型三角网格点或投影点为最近点,适合于毛坯件;特征域精配准则通过求解点云在平面/圆柱孔特征上的投影点为最近点,适合于成品件.试验及计算结果表明:全域配准的配准精度随表面离散点距离的减小而提高, 当离散点距离达到1.50 mm时,其配准精度已经达到0.15 mm,基本满足工程应用要求.当配准精度相同时,配准效率较其它方法提高10%~20%.      

基于分块EnKF的非线性目标跟踪算法

针对滤波航迹的相关性以及初始状态的选择会对跟踪性能产生影响的问题,将集合卡尔曼滤波算法引入到非线性目标跟踪领域,验证了其可行性和有效性,提出了基于分块集合卡尔曼滤波的非线性目标跟踪算法.采用分块思想生成初始集合,使用协方差矩阵加权方法解决分块间的航迹相关问题.仿真结果表明基于分块集合卡尔曼滤波的目标跟踪算法计算复杂度和以往的集合卡尔曼滤波算法同阶的情况下可以提供更高的运动参数估计精度,解决了粒子滤波算法计算量大难以进行实时跟踪的问题.      

基于激光与视觉融合的车辆自主定位与建图算法

定位与建图是车辆未知环境自主驾驶的基础,激光雷达依赖于场景几何特征而视觉图像 易受光线干扰,依靠单一激光点云或视觉图像的定位与建图算法存在一定局限性。本文提出一 种激光与视觉融合SLAM(Simultaneous Localization And Mapping)的车辆自主定位算法,通过融 合互补的激光与视觉各自优势提升定位算法的整体性能。为发挥多源融合优势,本文在算法前 端利用激光点云获取视觉特征的深度信息,将激光-视觉特征以松耦合的方式输入位姿估计模块 提升算法的鲁棒性。针对算法后端位姿和特征点大范围优化过程中计算量过大的问题,提出基 于关键帧和滑动窗口的平衡选取策略,以及基于特征点和位姿的分类优化策略减少计算量。实 验结果表明：本文算法的平均定位相对误差为 0.11 m 和 0.002 rad,平均资源占用率为 22.18% (CPU)和 21.50%(内存),与经典的 A-LOAM(Advanced implementation of LOAM)和 ORB-SLAM2 (Oriented FAST and Rotated BRIEF SLAM2)算法相比在精确性和鲁棒性上均有良好表现。     

Voxelizating 3D mesh models with gray levels

A novel algorithm to voxelize 3D mesh models with gray levels is presented in this paper. The key innovation of our method is to decide the gray level of a voxel according to the total area of all surfaces contained by it. During the preprocessing stage, a set of voxels in the extended bounding box of each triangle is established. Then we travel each triangle and compute the areas between it and its set of voxels one by one. Finally, each voxel is arranged a discrete gray level from 0 to 255. Experiments show that our algorithm gets a comparatively perfect result compared with the prevenient ones and approaches the original models in a more accurate way.     

带内外边界约束的平面点集Delaunay三角剖分

算法首先将离散点与约束边界点一起进行Delaunay三角剖分,形成初始Delaunay三角网,然后将约束边界上的各条约束线段通过局部更新依次嵌入已有的三角网,最后再删除多余的三角形,从而得到带内外边界约束的平面点集Delaunay三角剖分.     

一种新的基于LPMD的散射路径识别算法

NLOS（non-line-of-sight）非视距误差是地面无线定位的主要误差来源.为了从NLOS散射信号中识别出单次散射路径和多次散射路径,提出了一种新的基于LPMD（line-of-possible-mobile-device）目标可能位置线的识别算法.该算法以所有LPMD线的交点来计算初步参考点位置,并通过距离初步参考点较远的一部分交点对剩余较近的交点做负期望补偿,以修正参考点的位置,同时使用参考点到各LPMD线的直线距离来构造散射路径类型判决表达式,进行单/多次散射路径判决.在不同的仿真参数条件下,分别对普通场景和特殊场景进行了散射路径识别仿真实验,结果表明,当侦测站和目标点之间的距离达到4 km时,本文算法的虚警和漏警概率分别只有3%和9%,比原有基于LPMD算法的分别降低了21%和8%.      

带时间窗的多行程交换箱甩挂路径优化

针对市区-近郊的同城物流配送系统,为尽可能降低物流运输成本的同时提高客户满意度,研究带时间窗的具有多行程的交换箱甩挂运输问题. 在配送系统中,配送车辆从配送中心出发将货物运往各个客户点,由于道路条件的限制,客户点分为只允许小型卡车配送的限制点客户和卡车或带交换箱拖车的整车都可为其配送的灵活点客户. 在客户时间窗和多行程的约束条件下,建立以成本最小为优化目标的基于交换箱甩挂的路径优化模型,提出装箱算法与遗传算法混合的启发式求解算法. 算例验证了所提出算法的有效性,为带交换箱的甩挂车进行物流配送提供一定的决策指导和参考意见.     

基于机载LiDAR和Mean Shift算法提取森林密度

为获取森林密度信息,利用Mean Shift算法对森林点云进行单木分割提取森林密度信息.首先,以点云三维坐标和法向量作为特征向量,利用统计分析方法选择合适带宽及阈值,采用Mean Shift算法对点云进行初始分割;其次,对分割后的点云进行分析,加入灌木、杂草等过滤条件,得到树冠点云;然后,对树冠点云再次进行Mean Shift分割,并对每类树冠点云进行统计,以稳态点为粗略位置标记计算森林密度;最后,与地面实测数据进行验证.地面数据验证结果表明,平均计算精度达到90.0%以上,可满足林业应用需求;通过与分水岭法进行对比发现, Mean Shift方法获得的精度为92.5%,比分水岭方法70.0%高出22.5%,且避免了分水岭方法导致的过分割现象.      

基于层次集成学习的高光谱图像分类

为提高高光谱图像（HSI）分类精度,基于集成学习方法提出高光谱图像分类的层次集成学习新框架。采用两种集成学习策略：外部集成及内部集成。在外部集成阶段,构造多种高光谱图像的光谱和空间特征,使外部集成呈高度多样性,有利于提高分类精度;内部集成阶段,针对关联多特征集中的个体,Adaboost算法实现个体分类性能的提高。两组高光谱数据的实验结果表明,与原始的Adaboost和单分类器相比较,该方法在整体精度方面有更好的性能。     

一种改进的基于轮廓特征拐点的遮挡车辆分离方法

在传统的基于4类特征拐点的遮挡车辆分离方法的基础上进行改进,提出了一种基于8类特征拐点的分离方法．该方法以车辆常用的矩形模板为先验知识,首先对存在遮挡的连通区域提取边缘轮廓,并将轮廓上的特征拐点分为8类;然后在对相邻且同类的轮廓特征拐点进行合并的基础上,利用改进的车辆轮廓特征拐点的类型组合来实现遮挡车辆的识别和分离．仿真实验表明,本文所提出的新方法具有更好的鲁棒性和精确性,且方法简单,具有很高的实际应用价值．     

一般多边形窗口的有效线裁剪算法

在分析已有线裁剪算法优点和缺点的基础上，提出了一种新的一般多边形窗口线裁剪算法．这种新算法通过计算被裁剪线段和多边形各边的真实交点，利用交点将被裁剪线段分成若干子线段，用各子线段的中点来判断是否对它们实施裁剪．裁剪算法对各种情况都能快速、正确地得到结果．     

缓和曲线正交拟合的Levenberg-Marquardt算法

为了由测量点识别既有线路中的缓和曲线参数,研究了基于参数方程的缓和曲线正交拟合迭代优化方法.首先,通过特征值分析,阐明了由于病态性的存在,在迭代过程中,常规的Gauss-Newton (GN)算法会发散.其次,提出了双目标优化模型,将GN算法与最速下降法结合,确定了正交拟合缓和曲线的Levenberg-Marquardt(LM)算法.同时提出了在寻优过程中,评估当前迭代位置距离最优位置的远近来动态设置LM参数.最后以一段缓和曲线的实测点为例,随机取样了5 000例初值,采用蒙特卡罗方法对比了GN算法和LM算法拟合缓合曲线的性能.试验结果表明:GN算法拟合缓合曲线不收敛;对于不同的初始值,LM算法都收敛到相同的最优值,体现了LM算法具有良好的稳健性;LM算法的迭代次数最少为5次,最大为50次,平均为16.8次,迭代次数和初值与最优值位置的远近相关.     

基于K近邻和法向精度的点云精简算法

针对传统点云简化算法在精简散乱点云数据时经常丢失过多特征点的不足,提出了基于K近邻和法向精度的点云精简算法.该算法首先对输入的散乱点云数据建立K近邻索引,并剔除集群点及离群点,从而完成点云数据的预处理,然后对预处理后的数据进行Delaunay三角化,并重构三角网格面,最后依据法向精度进行非特征点剔除.仿真实验表明,该算法既能较大程度地精简点云数据,又能较好地保持原有模型的基本特征.     

遗传算法在逆向物流网络规划设计中的应用

考虑到客户、初始回收点和回收中心3个层次的逆向物流网络,选择合适的回收中心,确定初始回收点个数和回收周期,基于一些基本假设,以最小化总的相关成本建立了1个内嵌两个指派模型的非线性混合整数规划模型,对符号体系和求解思路进行了分析.借助遗传算法,结合具体实例,给出了求解步骤和求解结果分析.