基于密度聚类和投票判别的三维数据去噪方法

本文介绍的是一种有效发现和去除三维数据噪音方法,它既能改善去除噪音的效果,又能保持特征信息。该方法是一种两阶段噪音数据处理方法。该方法首先通过密度聚类将数据分类为正常簇集合、疑似簇集合、异常簇集合,然后利用正常簇集合中的点对疑似簇集合中各点进行投票判断,最终得到一个合理的三维点云数据模型。实验结果证明,本方法能够有效去除制造类工件模型的三维点云数据中的噪音数据,同时能良好保持模型表面的特征,加快处理效率。     

基于图像的重建点云离群点检测算法

基于图像重建出的三维点云模型通常会包含许多离群点,这些离群点可能孤立存在或密集聚集在一起形成点簇,也可能分布在模型周围甚至附着在模型表面。通过一种检测方法很难有效滤除多种分布状态的离群点,因此,提出了综合的离群点监测算法。首先通过空间距离剔除与模型主体较远的离群点,并通过构建空间拓扑关系加快离群点搜索速度;然后利用边界匹配法,将较小点簇分别与最大点簇进行对比,滤除模型周围离群点簇;最后采用改进的K-means算法,根据RGB颜色值特征对点云数据进行聚簇分类,结合已识别的离群点,检测和滤除附着在模型表面的离群点。仿真实验结果表明,此方法能够有效滤除点云模型中多种分布状态的离群点。     

散乱点云离群点的分类识别算法

散乱点云离群点识别和滤除是重建高质量曲面的前提,也是散乱点云预处理的重要步骤.提出一种散乱点云区域增长策略和一个基于曲面变化度的局部离群指标SVLOF,并将其应用到离群点识别中.通过分析离群点产生的原因,根据离群点到点云主体的距离将离群点分为远离群点和近离群点2类;对远离群点采用基于三维区域增长的方法进行识别,而对于近离群点采用SVLOF系数进行识别.基于仿真数据和实测数据的实验均表明,采用文中算法能够快速、有效地检测出孤立离群点和小型聚类离群点.     

基于数据场的类簇中心选取及其聚类

针对现有聚类算法普遍存在聚类质量低、参数依赖性大、孤立点难识别等问题,提出一种基于数据场的聚类算法。该算法通过计算每个数据对象点的势值,根据类簇中心的势值比周围邻居的势值大,且与其他类簇中心有相对较大距离的特点,确定类簇中心;根据孤立点的势值等于零的特点,选出孤立点;最后将其他数据对象点划分到比自身势值大且最近邻的类簇中,从而实现聚类。仿真实验表明,该算法在不需要人为调参的情况下准确找出类簇中心和孤立点,聚类效果优良,且与数据集的形状无关。     

Spectral moving removal of non-isolated surface outlier clusters

In this paper, we propose a new algorithm for the fast removal of non-isolated surface outlier clusters. It consists of three basic components: (a) an intrinsic metric for detecting outliers on the basis of minimum variance principle; (b) bi-means clustering of a normalized histogram; (c) surface propagation for a geometric coherence check. The unique contributions of our approach include (a) a new idea of identifying non-isolated outlier clusters and linking the local spectral property to a global outlier removal process; (b) a modified data clustering scheme with a geometric coherence check. In comparison with existing algorithms, our algorithm is evaluated in terms of the quality and computational cost of outlier removal. Numerical experiments indicate the effectiveness of our approach in the aspects of convergence, accuracy, time and space efficiency.     

An adaptive normal estimation method for scanned point clouds with sharp features

Normal estimation is an essential task for scanned point clouds in various CAD/CAM applications. Many existing methods are unable to reliably estimate normals for points around sharp features since the neighborhood employed for the normal estimation would enclose points belonging to different surface patches across the sharp feature. To address this challenging issue, a robust normal estimation method is developed in order to effectively establish a proper neighborhood for each point in the scanned point cloud. In particular, for a point near sharp features, an anisotropic neighborhood is formed to only enclose neighboring points located on the same surface patch as the point. Neighboring points on the other surface patches are discarded. The developed method has been demonstrated to be robust towards noise and outliers in the scanned point cloud and capable of dealing with sparse point clouds. Some parameters are involved in the developed method. An automatic procedure is devised to adaptively evaluate the values of these parameters according to the varying local geometry. Numerous case studies using both synthetic and measured point cloud data have been carried out to compare the reliability and robustness of the proposed method against various existing methods.     

基于数学形态学的模糊异常点检测

异常点检测作为数据挖掘的一项重要任务,可能会导致意想不到的知识发现.但传统的异常点检测技术都忽略了数据的自然结构,即异常点与簇的联系.然而,把异常点得分和聚类方法结合起来有利于对异常点与簇的联系的研究.提出基于数学形态学的模糊异常点检测与分析,把数学形态学技术和基于连接的异常点检测方法集成到一个模糊模型中,从异常隶属度和模糊隶属度这两个方面来分析对象与簇集的模糊关系.通过充分的实验证明,该算法能够对复杂面状和变密度的数据集,正确、高效地找出异常点,同时发现与异常点相关联的簇信息,探索异常点与簇核的关联深度,对异常点本身的意义具有启发作用.     

基于近邻传播的离群点检测算法

离群点是与其他正常点属性不同的一类对象,其检测技术在各行业上均有维护数据纯度、保障业内安全等重要应用,现有算法大多是基于距离、密度等传统方法判断检测离群点.本算法给每个对象分配一个"孤立度",即该点相对其邻点的孤立程度,通过排序进行判定,比传统算法效率更高.在AP(affinity propagation)聚类算法的基础上进行改进与优化,提出能检测异常数据点的算法APO(outlier detection algorithm based on affinity propagation).通过加入孤立度模块并计算处理样本点的孤立信息,并引入放大因子,使其与正常点之间的差异更明显,通过增大算法对离群点的敏感性,提高算法的准确性.分别在模拟数据集和真实数据集上进行对比实验,结果表明:该算法与AP算法相比,对离群点的敏感性更加强烈,且本算法检测离群点的同时也能聚类,是其他检测算法所不具备的.     

散乱点云近离群点识别算法

针对原始曲面变化度的局部离群系数(SVLOF)无法有效滤除三维实体的棱边或棱角处的离群点问题,提出了一种散乱点云近离群点的滤除算法。该算法首先将SVLOF定义在类k邻域上,并将SVLOF的定义内容进行了扩展,使其既能滤除平滑曲面上的离群点,又能滤除三维实体的棱边或棱角点处的离群点,同时仍然保留SVLOF原有的足够宽泛的阈值选取空间。仿真数据和实际数据的实验结果均表明,在效率基本保持不变的情况下,所提算法能比原始SVLOF算法更有效地检测出距离主体点云近的离群点。     

An extension on robust directed projection of points onto point clouds

Azariadis and Sapidis [Azariadis PN, Sapidis NS. Drawing curves onto a cloud of points for point-based modelling. Computer-Aided Design 2005;37(1):109-22] introduced a novel method of point directed projection (DP) onto a point cloud along an associated projection vector. This method is essentially based on an idea of least sum of squares by making use of a weight function for bounding the influence of noise. One problem with their method is the lack of robustness for outliers. Here, we present a simple, robust, and efficient algorithm: robust directed projection (RDP) to guide the DP computation. Our algorithm is based on a robust statistical method for outlier detection: least median of squares (LMS). In order to effectively approximate the LMS optimization, the forward search technique is utilized. The algorithm presented here is better suited to detect outliers than the DP approach and thus finds better projection points onto the point cloud. One of the advantages of our algorithm is that it automatically ignores outliers during the directed projection phase.     

Multi-scale tensor voting for feature extraction from unstructured point clouds

Identifying sharp features in a 3D model is essential for shape analysis, matching and a wide range of geometry processing applications. This paper presents a new method based on the tensor voting theory to extract sharp features from an unstructured point cloud which may contain random noise, outliers and artifacts. Our method first takes the voting tensors at every point using the corresponding neighborhoods and computes the feature weight to infer the local structure via eigenvalue analysis of the tensor. The optimal scale for a point is automatically determined by observing the feature weight variation in order to deal with both a noisy smooth region and a sharp edge. We finally extract the points at sharp features using adaptive thresholding of the feature weight and the feature completion process. The multi-scale tensor voting of a given point set improves noise sensitivity and scale dependency of an input model. We demonstrate the strength of the proposed method in terms of efficiency and robustness by comparing it with other feature detection algorithms.     

Delaunay三角剖分在离群点检测中的应用

在传统的基于[K]近邻的算法中,需要为算法设置邻居参数[k]的值,只有具备相关的先验知识才能确定合适的参数值。为了减少参数对于离群点检测的影响,提出了一种无需参数的基于Delaunay三角剖分的离群点检测算法。Delaunay三角剖分是数值分析以及图形学中的重要基础理论,它的构建无需任何参数,在三角剖分图中的每个数据对象与它空间上相邻的点都存在边直接相连,因此可以形成一种有效的邻居关系。算法首先通过Delaunay三角剖分形成每个点的空间邻居集合,然后根据每个点与它们空间邻居之间的分布特征,计算它们的离群程度,根据离群程度的大小判断该点是否为离群点。通过实验与相关的算法比较,算法具有更好的效果。     

基于深度学习的道路障碍物检测方法

针对基于激光雷达（LiDAR）的三维点云数据处理及道路障碍物检测的问题,提出一种基于深度学习的路障碍物检测方法。首先,采用统计滤波算法对原始点云进行离群点的剔除处理;其次,提出一种端到端的深度神经网络VNMax,利用最大池化对区域候选网络（RPN）架构进行优化,构建改进的目标检测层;最后,在KITTI数据集上进行了训练及测试实验。结果显示,经过滤波处理,点云中各点之间的平均距离得到有效减少。通过对在KITTI数据集的简单、中等和困难任务的车辆定位处理结果比较得出,所提方法的平均精度比VoxelNet（Unofficial）分别提高了11.3个百分点、6.02个百分点和3.89个百分点。实验测试结果表明,统计滤波算法仍是有效的三维点云数据处理手段,最大池化模块可以提高深度神经网络的学习性能和目标定位能力。     

SIFT算法在点云配准中的应用

提出一种精确有效的点云配准算法。通过对图像进行SIFT特征检测与匹配来获得特征点与匹配关系,用RANSAC算法剔除误匹配点,然后由映射关系获取三维对应特征点,采用投票法来进一步剔除误匹配点。在由单位四元数法获得点云初始位置关系的基础上,采用基于特征点的改进ICP算法来实现精确配准。通过实验验证,该算法在点云配准中具有速度快和稳定性好的特点。     

ODIC-DBSCAN:一种新的簇内孤立点分析算法

长期以来,孤立点的检测一直聚焦于簇边缘的离散点,当聚类后簇的数目低于实际数目,或孤立点被伪装在簇内的情况下,簇内孤立点的判定则会更加困难.为判定簇内孤立点,提出一种基于密度聚类DBSCAN（Density based spatial clustering of application with noise）的簇内孤立点检测方法ODIC-DBSCAN（Outlier detection of inner-cluster based on DBSCAN）.首先在建立距离矩阵的基础上,通过半径获取策略得到针对该点集的k个有效半径Radius集合,并据此构造密度矩阵;然后建立点集覆盖模型,提出了相邻有效半径构造的覆盖多维体能够覆盖点集的思想,并通过拉格朗日乘子法求取最优的覆盖多维体数目之比,输出点比阈值组;最后重建ODIC-DBSCAN的孤立点检测方法,以簇发生融合现象作为算法终止的判定条件.实验通过模拟数据集,公开benchmark与UCI数据集共同验证了ODIC-DBSCAN算法,展示了聚类过程;分析了算法性能;并通过与其他聚类、孤立点判定方法的对比,验证了算法对簇内孤立点的判定效果.     

A fuzzy index for detecting spatiotemporal outliers

The detection of spatial outliers helps extract important and valuable information from large spatial datasets. Most of the existing work in outlier detection views the condition of being an outlier as a binary property. However, for many scenarios, it is more meaningful to assign a degree of being an outlier to each object. The temporal dimension should also be taken into consideration. In this paper, we formally introduce a new notion of spatial outliers. We discuss the spatiotemporal outlier detection problem, and we design a methodology to discover these outliers effectively. We introduce a new index called the fuzzy outlier index, FoI, which expresses the degree to which a spatial object belongs to a spatiotemporal neighbourhood. The proposed outlier detection method can be applied to phenomena evolving over time, such as moving objects, pedestrian modelling or credit card fraud.     

应用于三维点云数据去噪的改进C均值算法

针对三维激光扫描仪采集到的点云数据中离群点不易区分和去噪难度大的问题,提出了一种改进的C均值算法。通过分析三维点云数据特征,在传统C均值算法中引入模糊聚类权重因子,降低类内距离和拉大类间距离,有效增强了离群点特征以降低识别难度。进而将识别出的噪声分类别处理,利用改进的C均值算法去除大尺度噪声,构造双边滤波算法去除小尺度噪声数据。与密度聚类算法、正交整体最小二乘平面拟合和基于特征选择的双边滤波点云去噪等算法相比,去噪准确度分别提升了7.3%、6.5%和6.0%,实验结果表明该算法可以有效去除大尺度噪声并能较好地保留有效数据。     

伪点云修正增强激光雷达数据

目的 激光雷达在自动驾驶中具有重要意义,但其价格昂贵,且产生的激光线束数量仍然较少,造成采集的点云密度较稀疏。为了更好地感知周围环境,本文提出一种激光雷达数据增强算法,由双目图像生成伪点云并对伪点云进行坐标修正,进而实现激光雷达点云的稠密化处理,提高3D目标检测精度。此算法不针对特定的3D目标检测网络结构,是一种通用的点云稠密化方法。方法 首先利用双目RGB图像生成深度图像,根据先验的相机参数和深度信息计算出每个像素点在雷达坐标系下的粗略3维坐标,即伪点云。为了更好地分割地面,本文提出了循环RANSAC （random sample consensus）算法,引入了一个分离平面型非地面点云的暂存器,改进复杂场景下的地面分割效果。然后将原始点云进行地面分割后插入KDTree （k-dimensional tree）,以伪点云中的每个点为中心在KDTree中搜索若干近邻点,基于这些近邻点进行曲面重建。根据曲面重建结果,设计一种计算几何方法导出伪点云修正后的精确坐标。最后,将修正后的伪点云与原始激光雷达点云融合得到稠密化点云。结果 实验结果表明,稠密化的点云在视觉上具有较好的质量,物体具有更加完整的形状和轮廓,并且在KITTI （Karlsruhe Institute of Technology and Toyota Technological Institute）数据集上提升了3D目标检测精度。在使用该数据增强方法后,KITTI数据集下AVOD （aggregate view object detection）检测方法的AP_3D-Easy （average precision of 3D object detection on easy setting）提升了8.25%,AVOD-FPN （aggregate view object detection with feature pyramid network）检测方法的AP_BEV-Hard （average precision of bird’s eye view on hard setting）提升了7.14%。结论 本文提出的激光雷达数据增强算法,实现了点云的稠密化处理,并使3D目标检测结果更加精确。     

基于聚类的两段式孤立点检测算法

现有的大多数孤立点检测算法都需要预先设定孤立点个数,并且还缺乏对不均匀数据集的检测能力。针对以上问题,提出了基于聚类的两段式孤立点检测算法,该算法首先用DBSCAN聚类算法产生可疑孤立点集合,然后利用剪枝策略对数据集进行剪枝,并用基于改进距离的孤立点检测算法产生最可能孤立点排序集合,最终由两个集合的交集确定孤立点集合。该算法不必预先设定孤立点个数,具有较高的准确率与检测效率,并且对数据集的分布状况不敏感。数据集上的实验结果表明,该算法能够高效、准确地识别孤立点。     

基于划分的高效异常轨迹检测

为了在海量轨迹数据库中高效准确地挖掘出异常轨迹,提出了基于划分的异常轨迹检测算法。该算法通过计算局部轨迹点之间的匹配程度来探测异常轨迹,将异常轨迹检测由形状匹配问题转化为传统的异常点检测问题,并设计了一种基于空间划分的网格索引结构,提高算法的运行效率。实验证明,该算法不仅具有较高的挖掘效率,而且能够检测出更具实际意义的异常轨迹。