零件点云法向量估计的多尺度特征融合网络

为了解决机械零配件点云处理中非均匀采样干扰、尖锐特征损失等难点,提出一种基于深度神经网络多尺度融合的点云法向量估计方法.该网络在不同邻域尺度下集成了采样点细节与点云块整体两种特征.为了使该多维回归输出网络的训练更稳定且能缓解梯度爆炸问题,重新设计了一个光滑的损失函数.实验结果表明,该方法性能优于传统的方法以及HoughCNN、PCPNet等方法,能够更准确地估计尖锐边缘的法向量,对点云各种噪声和采样方法鲁棒性都更强.     

向量相似度可复原三维点云压缩算法

针对三维点云数据压缩中细节特征不易保留,模型平缓部位存在过度压缩以及压缩后的点云模型不易复原等问题,提出一种基于向量相似度的三维点云压缩算法和复原算法CVS。向量相似性度量采用提出的L3A进行度量。CVS把每个三维坐标点看作是连接其坐标和原点的三维向量,按照三维坐标点的读入顺序选取参考向量,生成覆盖整个点云区域的采样区域,进行分区压缩。在采样区域中使用最小二乘曲面拟合算法对包含其中的点云进行曲面拟合,设置曲率阈值剔除坐标点,并存储曲面方程参数用于复原。通过控制L3A向量相似度中的长度和角度的变化阈值,使得密集点云区域的压缩率高于非密集区域的压缩率,通过控制曲率阈值,使得低曲率区域的压缩率高于高曲率区域的压缩率,最大程度保留模型细节特征。CVS使用压缩阶段产生的复原信息生成点云来恢复模型的细节特征,使得模型特征更加明显。     

基于深度学习的三维点云修复技术综述

三维点云是最常用的三维场景/物体表示方法之一.根据点云修复侧重点不同,将基于深度学习的三维点云修复技术划分为密集重建、补全重建和去噪重建3类;详细分析了相关典型修复模型及关键技术,如特征编码、特征扩展和损失函数设计;总结了常用的网络模块、点云数据集和评估准则;最后讨论了3类修复技术之间的关系,并从旋转不变性特征提取、细节信息修复、拓扑关系保持、几何算法应用和多模态数据融合5个方面探讨了点云修复技术面临的挑战及未来发展趋势.     

散乱点云数据曲率估计方法

针对带有强噪声离散点云数据曲率计算问题,提出一种基于稳健统计的曲率估计方法。首先,用一个二次曲面拟合三维空间采样点处的局部形状;其次,随机地选择该采样点邻域内的子集,多次执行这样的拟合过程,通过变窗宽的最大核密度估计,就得到了最优拟合曲面;最后,将采样点投影到该曲面上,计算投影点曲率信息,就得到采样点曲率。实验结果表明,所提方法对噪声和离群点是稳健的,特别是随着噪声方差的增大,要明显好于传统的抛物拟合方法。     

利用Kinect深度信息的三维点云配准方法研究

针对三维重建中的点云配准问题,提出一种基于点云特征的自动配准算法。利用微软Kinect传感器采集物体的多视角深度图像,提取目标区域并转化为三维点云。对点云进行滤波并估计快速点特征直方图特征,结合双向快速近似最近邻搜索算法得到初始对应点集,并使用随机采样一致性算法确定最终对应点集。根据奇异值分解法求出点云的变换矩阵初始值,在初始配准的基础上运用迭代最近点算法做精细配准。实验结果表明,该配准方法既保证了三维点云的配准质量,又降低了计算复杂度,具有较高的可操作性和鲁棒性。     

规则特征曲面点云法向估计

目的 针对特征曲面点云法矢估计不准确,点云处理时容易丢失曲面的细节特征等问题,提出基于高斯映射的特征曲面散乱点云法向估计法。方法 首先,用主成分分析法粗略地估算点云法向和特征点;其次,将特征点的各向同性邻域映射到高斯球,用K均值聚类法对高斯球上的数据分割成多个子集,以最优子集对应的各向异性邻域拟合曲面来精确估算特征点的法向量;最后,通过测试估计法向与标准法向的误差来评价估计法矢的准确性,并且将估计的法向应用到点云曲面重建中来比较特征保留效果。结果 本文方法估计的法向最小误差接近0,对噪声有较好的鲁棒性,重建的曲面能保留曲面的尖锐特征,相比于其他法向估计法,所提出的方法估计的法向更准确。结论 本文方法能够比较准确的估算尖锐特征曲面法向量,对噪声鲁棒性强,具有较高的适用性。     

基于特征点动态选择的三维人脸点云模型重建

针对典型的点云配准方法中伪特征点过多导致配准效率低和配准结果不精确的问题,提出一种基于特征点动态选择的三维人脸点云模型重建方法。该方法在粗配准阶段,采用动态特征矩阵求解法获取粗匹配特征变换矩阵以避免伪特征点的干扰。在精配准过程中,采用二次加权法向量垂直距离法在人脸流形表面选择更有效的特征点以减少伪特征点的数量,并采用基于特征融合与局部特征一致性的迭代最近点方法进行精配准。经过对比实验验证了算法的可行性,实验结果表明,提出算法能够实现高精度且快速的三维人脸点云模型重建,且均方根误差达到1.816 5 mm,相较于其他算法,在模型重建精度和效率方面都有所提升,具有良好的应用前景。     

基于随机分布估计的点云密度提取

针对目前密度提取方法提取的密度信息不能表现点云局部分布信息和分布随机性的缺陷,提出结合随机分布估计的密度提取方法。该方法采用分块计数法得到每个小分块的密度,结合点云总体的密集度得到一个能够反映点云局部积聚特征的参数,为判另Ⅱ点云分布的随机性、均匀性等提供较好的特征依据。     

基于改进法向量估计的多视角点云配准算法

针对多视角三维测量中多片点云重叠区域提取及高精度配准的问题,本文提出一种多视角异源低重叠率点云配准方法。首先基于点云之间的初始位置,互相计算源点云和目标点云彼此的最近点集,自动提取两片点云重叠部分;然后使用迭代最近点算法精配准重叠点云。通过法向量特征进一步提高点云配准精度,并提出改进点云法向量估计算法用以剔除错误匹配点对,显著减小了复杂结构点云配准的距离均方根误差。结果表明,使用经典点云数据仿真实验验证了该算法的性能,并通过多视角条纹投影三维测量系统采集点云数据验证了算法的有效性。     

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.     

Robust normal estimation and region growing segmentation of infrastructure 3D point cloud models

Modern remote sensing technologies such as three-dimensional (3D) laser scanners and image-based 3D scene reconstruction are in increasing demand for applications in civil infrastructure design, maintenance, operation, and as-built construction verification. The complex nature of the 3D point clouds these technologies generate, as well as the often massive scale of the 3D data, make it inefficient and time consuming to manually analyze and manipulate point clouds, and highlights the need for automated analysis techniques. This paper presents one such technique, a new region growing algorithm for the automated segmentation of both planar and non-planar surfaces in point clouds. A core component of the algorithm is a new point normal estimation method, an essential task for many point cloud processing algorithms. The newly developed estimation method utilizes robust multivariate statistical outlier analysis for reliable normal estimation in complex 3D models, considering that these models often contain regions of varying surface roughness, a mixture of high curvature and low curvature regions, and sharp features. An adaptation of Mahalanobis distance, in which the mean vector and covariance matrix are derived from a high-breakdown multivariate location and scale estimator called Deterministic MM-estimator (DetMM) is used to find and discard outlier points prior to estimating the best local tangent plane around any point in a cloud. This approach is capable of more accurately estimating point normals located in highly curved regions or near sharp features. Thereafter, the estimated point normals serve a region growing segmentation algorithm that only requires a single input parameter, an improvement over existing methods which typically require two control parameters. The reliability and robustness of the normal estimation subroutine was compared against well-known normal estimation methods including the Minimum Volume Ellipsoid (MVE) and Minimum Covariance Determinant (MCD) estimators, along with Maximum Likelihood Sample Consensus (MLESAC). The overall region growing segmentation algorithm was then experimentally validated on several challenging 3D point clouds of real-world infrastructure systems. The results indicate that the developed approach performs more accurately and robustly in comparison with conventional region growing methods, particularly in the presence of sharp features, outliers and noise.     

3D人耳点云配准的并行Softassign算法

提出了一种新的人耳点云并行Softassign配准算法.在基于CUDA对Softassign算法进行并行加速的基础上,利用三维点云离散曲率估计和三维空间kd-tree相结合的方法,对三维人耳点云进行点云简化,使简化后人耳点云能够保留足够的几何特征,然后对简化人耳点云进行Softassign配准,提高Softassign算法在人耳点云配准中的配准精度,从而避免了局部配准等缺陷,并在实际应用中验证了算法效率和精度.     

结合图像分割和点云分割的障碍物检测算法

针对单目图像检测障碍物的低可靠性和当前双目视觉检测障碍物的局限性的问题,提出一种结合图像分割和点云分割技术的双目视觉障碍物检测方法。通过设定检测深度范围,分割障碍物点云与道路点云;采用将分割出的障碍物点云对应的视差图与图像分割得到的子图进行比较的策略,有效解决对不同深度、倾斜面和不规则障碍物检测效果差的问题。通过实验验证了在获得稀疏三维点云的情况下,该方法对障碍物的检测具有较好的鲁棒性。     

基于曲率的点云数据配准算法

为了实现不同视角下测得的数据的多视定位,提出一种点云数据配准算法。该算法针对近邻内的点,采用二次曲面逼近的方法来求得每个点的曲率,并根据曲率的Hausdorff距离来寻找有效点集,建立名义上的对应关系,最后用四元组法来求得坐标变换,把数据统一到一个坐标系下。该算法利用曲率的性质准确判断对应点集,解决了任意多视点云的拼合问题,试验结果验证了其有效性和精度。     

Collision detection between point clouds using an efficient k-d tree implementation

Context: An important task in civil engineering is the detection of collisions of a 3D model with an environment representation. Existing methods using the structure gauge provide an insufficient measure because the model either rotates or because the trajectory makes tight turns through narrow passages. This is the case in either automotive assembly lines or in narrow train tunnels.Objective: Given two point clouds, one of the environment and one of a model and a trajectory with six degrees of freedom along which the model moves through the environment, find all colliding points of the environment with the model within a certain clearance radius.Method: This paper presents two collision detection (CD) methods called kd-CD and kd-CD-simple and two penetration depth (PD) calculation methods called kd-PD and kd-PD-fast. All four methods are based on searches in a k-d tree representation of the environment. The creation of the k-d tree, its search methods and other features will be explained in the scope of their use to detect collisions and calculate depths of penetration.Results: The algorithms are benchmarked by moving the point cloud of a train wagon with 2.5 million points along the point cloud of a 1144 m long train track through a narrow tunnel with overall 18.92 million points. Points where the wagon collides with the tunnel wall are visually highlighted with their penetration depth. With a safety margin of 5 cm kd-PD-simple finds all colliding points on its trajectory which is sampled into 19,392 positions in 77 s on a standard desktop machine of 1.6 GHz.Conclusion: The presented methods for collision detection and penetration depth calculation are shown to solve problems for which the structure gauge is an insufficient measure. The underlying k-d tree is shown to be an effective data structure for the required look-up operations.     

点云模型法矢调整优化算法

点云中存在奇异情况时,采用最小生成树法进行法矢调整会出现错误,而采用曲面重建方法运算效率又较低,为此提出一种点云模型法矢调整的优化算法.算法分别处理薄壁特征、垂直法向和相邻曲面3种奇异情况.对薄壁特征,算法提取特征点并在该处强制进行法矢取反;对垂直法向,算法通过扩大邻域搜索范围来获得法矢变化趋势;对相邻曲面,算法在K邻域中剔除歧义邻域点,避免在最小生成树中生成错误边.实验结果表明,该算法在点云中存在奇异情况时能够进行正确的法矢调整,并且相较于曲面重建方法具有较高的效率.     

Multi-contour initial pose estimation for 3D registration

Reliable manipulation of everyday household objects is essential to the success of service robots. In order to accurately manipulate these objects, robots need to know objects’ full 6-DOF pose, which is challenging due to sensor noise, clutters, and occlusions. In this paper, we present a new approach for effectively guessing the object pose given an observation of just a small patch of the object, by leveraging the fact that many household objects can only keep stable on a planar surface under a small set of poses. In particular, for each stable pose of an object, we slice the object with horizontal planes and extract multiple cross-section 2D contours. The pose estimation is then reduced to find a stable pose whose contour matches best with that of the sensor data, and this can be solved efficiently by cross-correlation. Experiments on the manipulation tasks in the DARPA Robotics Challenge validate our approach. In addition, we also investigate our method’s performance on object recognition tasks raising in the challenge.