3D Point Cloud Registration Research Articles

A preprocessing method of 3D point clouds registration in urban environments

Aiming at the problem that 3D LiDAR point cloud has high data density, outlier noise, and scattered distribution in urban environment, which is not conducive to the matching between point clouds in the later stage, a pre-processing method for large-scale LiDAR point cloud frame matching in urban environments is proposed. First, the point cloud data is transformed into a Mean Elevation Map, and the ground point segmentation processing is performed on the point cloud using the height gradient between the grids; then, the DBSCAN clustering algorithm is improved by the three-dimensional voxel grid division method, and the improved VG-DBSCAN is used to cluster point clouds and separate the target point cloud from the outliers after clustering, thereby, which eliminates outlier noises in the point cloud. Finally, the Voxel Grid filter is used to down sample the point cloud. The experimental results show that the proposed method can perform real-time preprocessing on point cloud data, and the average time is 132.1 ms. After pre-processing, the accuracy of point cloud frame matching is increased by 2 times, and the average time consumption is only 1/6 before pre-processing.

GridDS: a hybrid data structure for residue computation in point set matching

Registration of 3D point clouds is a problem that arises in a variety of research areas such as computer vision, computer graphics and computational geometry. This situation causes most papers in the area to focus on solving practical problems by using data structures often developed in theoretical contexts. Consequently, discrepancies arise between asymptotic cost and experimental performance. The point cloud registration or matching problem encompasses many different steps. Among them, the computation of the distance between two point sets (often refereed to as residue computation) is crucial and can be seen as an aggregate of range searching or nearest neighbor searching. In this paper, we aim at providing theoretical analysis and experimental performance of range searching and nearest neighbor data structures applied to 3D point cloud registration. Performance of widely used data structures such as compressed octrees, KDtrees, BDtrees and regular grids is reported. Additionally, we present a new hybrid data structure named GridDS, which combines a regular grid with some preexisting “inner” data structure in order to meet the best asymptotic bounds while also obtaining the best performance. The experimental evaluation in both synthetic and real data demonstrates that the hybrid data structures built using GridDS improve the running times of the single data structures. Thus, as we have studied the performances of the state-of-the-art techniques managing to improve their respective running times thanks to GridDS, this paper presents the best running time for point cloud residue computation up to date.

MONITORING OF PROGRESSIVE DAMAGE IN BUILDINGS USING LASER SCAN DATA

Abstract. Vulnerability of buildings to natural and man-induced hazards has become a main concern for our society. Ensuring their serviceability, safety and sustainability is of vital importance and the main reason for setting up monitoring systems to detect damages at an early stage. In this work, a method is presented for detecting changes from laser scan data, where no registration between different epochs is needed. To show the potential of the method, a case study of a laboratory test carried out at the Stevin laboratory of Delft University of Technology was selected. The case study was a quasi-static cyclic pushover test on a two-story high unreinforced masonry structure designed to simulate damage evolution caused by cyclic loading. During the various phases, we analysed the behaviour of the masonry walls by monitoring the deformation of each masonry unit. First a plane is fitted to the selected wall point cloud, consisting of one single terrestrial laser scan, using Principal Component Analysis (PCA). Second, the segmentation of individual elements is performed. Then deformations with respect to this plane model, for each epoch and specific element, are determined by computing their corresponding rotation and cloud-to-plane distances. The validation of the changes detected within this approach is done by comparison with traditional deformation analysis based on co-registered TLS point clouds between two or more epochs of building measurements. Initial results show that the sketched methodology is indeed able to detect changes at the mm level while avoiding 3D point cloud registration, which is a main issue in computer vision and remote sensing.

Fast registration of 3D point clouds with offset surfaces in precision grinding of free-form surfaces

Because of its high sensitivity to misalignment, precision grinding of free-form surfaces with micron accuracy requires accurate registration of the surface measurement point cloud. Registration of point clouds obtained with a coordinate measuring machine (CMM) is generally an iterative process of finding optimal coordinate transformation between the CMM frame and the model frame of the workpiece by minimizing the point-to-surface distances with probe radius compensation. For free-form surfaces, frequent calculation of point-to-surface distances consumes very much time, and a trade-off has to be made between the efficiency and the accuracy. This paper presents a method for fast registration of free-form surface point clouds based on the point-to-triangle distance which involves only Delaunay triangulation of a two-dimensional dataset, and the surface normal is quickly calculated from cross product. Probe radius compensation is realized by registering the probe center points with the offset surface. We prove that it is equivalent to registering the probe contact points with the nominal surface through theoretical analysis. The registration problem is then formulated as sequential linear least-square problems with properly defined ball constraints. To validate the method, numerical simulations are presented to show the accuracy of the point-to-triangle distance. The registration algorithm also shows excellent robustness against misalignment of tens of millimeters/degrees. Finally measurement, registration, and grinding of a free-form optical surface are experimentally demonstrated. The surface error obtained after registration is used for compensatory grinding which reduces it to micron level.

Robust 3D point cloud registration based on bidirectional Maximum Correntropy Criterion.

This paper presents a robust 3D point cloud registration algorithm based on bidirectional Maximum Correntropy Criterion (MCC). Comparing with traditional registration algorithm based on the mean square error (MSE), using the MCC is superior in dealing with complex registration problem with non-Gaussian noise and large outliers. Since the MCC is considered as a probability measure which weights the corresponding points for registration, the noisy points are penalized. Moreover, we propose to use bidirectional measures which can maximum the overlapping parts and avoid the registration result being trapped into a local minimum. Both of these strategies can better apply the information theory method to the point cloud registration problem, making the algorithm more robust. In the process of implementation, we integrate the fixed-point optimization technique based on the iterative closest point algorithm, resulting in the correspondence and transformation parameters that are solved iteratively. The comparison experiments under noisy conditions with related algorithms have demonstrated good performance of the proposed algorithm.

Compression and registration of 3D point clouds using GMMs

3D data sensors provide an enormous amount of information. It is necessary to develop efficient methods to manage this information under certain time, bandwidth or storage space requirements. In this work, we propose a 3D compression and decompression method. This method also allows the use of the compressed data for a registration process. First, points are selected and grouped, using a 3D-model based on planar surfaces. Next, we use a fast variant of Gaussian Mixture Models and an Expectation-Maximization algorithm to replace the points grouped in the previous step with a set of Gaussian distributions. These learned models can be used as features to find matches between two consecutive poses and apply 3D pose registration using RANSAC. Finally, the 3D map can be obtained by decompressing the models.

Local voxelized structure for 3D binary feature representation and robust registration of point clouds from low-cost sensors

Local feature-based 3D point cloud registration is a central issue in the fields of 3D computer vision and robotics, and most previously proposed 3D local features are real-valued. In this paper (1) a novel binary descriptor named local voxelized structure (LoVS) for 3D local shape description and (2) a LoVS-based registration algorithm for low-quality, e.g., Kinect-captured, point clouds are proposed. LoVS simply encodes the local shape structure represented by point clouds into bit string using point spatial locations without computing complex geometric attributes, such as normals and curvature, at the feature representation stage. Specifically, the LoVS descriptor is extracted within a local cubic volume around the keypoint. The orientation of the cubic volume is determined by a local reference frame (LRF) to achieve rotation invariance. Then, the cubic volume is uniformly split into a set of voxels. A voxel is labeled 1 if it contains points; otherwise, 0. All the labels are integrated into the LoVS descriptor. Based on the LoVS descriptor, a robust and accurate point cloud registration algorithm was developed, which effectively handles the challenges presented by low-cost sensors, e.g., noise and varying data resolutions. Experiments and extensive comparisons on three descriptor-matching benchmarks and a large-scale Kinect point cloud registration dataset show the effectiveness and the over-all superiority of our proposed LoVS descriptor and LoVS-based point cloud registration algorithm.

SLAM-driven robotic mapping and registration of 3D point clouds

Abstract With the rapid advancement of laser scanning and photogrammetry technologies, frequent geometric data collection at construction sites by contractors has been increased for the purpose of improving constructability, productivity, and onsite safety. However, the conventional static laser scanning method suffers from operational limitations due to the presence of many occlusions commonly found in a typical construction site. Obtaining a complete scan of a construction site without information loss requires that laser scans are obtained from multiple scanning locations around the site, which also necessitates extra work for registering each scanned point cloud. As an alternate solution to this problem, this paper introduces an autonomous mobile robot which navigates a scan site based on a continuously updated point cloud map. This mobile robot system utilizes the 2D Hector Simultaneous Localization and Mapping (SLAM) technique to estimate real-time positions and orientations of the robot in the x-y plane. Then, the 2D localization information is used to create 3D point clouds of unknown environments in real time to determine its navigation paths as a pre-scanning process. The advantage of this framework is the ability to determine the optimal scan position and scan angle to reduce the scanning time and effort for gathering high resolution point cloud data in real-time. The mobile robot system is able to capture survey-quality RGB-mapped point cloud data, and automatically register the scans for geometric reconstruction of the site. The performance of the overall system was tested in an indoor environment and validated with promising results.

4-Plane congruent sets for automatic registration of as-is 3D point clouds with 3D BIM models

Abstract Construction quality and progress control are demanding, yet critical construction activities. Building Information Models and as-built scanned data can be used in Scan-vs-BIM processes to effectively and comprehensively support these activities. This however requires accurate registration of scanned point clouds with 3D (BIM) models. Automating such registration remains a challenge in the context of the built environment, because as-built can be incomplete and/or contain data from non-model objects, and construction buildings and other structures often present symmetries and self-similarities that are very challenging to registration. In this paper, we present a novel automatic coarse registration method that is an adaptation of the ‘4 Points Congruent Set’ algorithm to the use of planes; we call it the ‘4-Plane Congruent Set’ (4-PlCS) algorithm. The approach is further integrated in a software system that delivers not one but a ranked list of the most likely transformations, so to allow the user to quickly select the correct transformation, if need be. Two variants of the method are also considered, in particular one in the case when the vertical axis is known a priori; we call that method the 4.5-PlCS method. The proposed algorithm is tested using five different datasets, including three simulated and two real-life ones. The results show the effectiveness of the proposed method, where the correct transformation always ranks very high (in our experiments, first or second), and is extremely close to the ground-truth transformation. Experimental comparison of the proposed approach with a standard, more intuitive approach based on finding 3-plane congruent sets shows the discriminatory power of 4-plane bases over 3-plane bases, albeit at no clear benefits in terms of computational time. The experimental results for the 4.5-PlCS method show that it delivers a non-negligible reduction in computational time (approx. 20%), but at no additional benefit in terms of effectiveness in finding the correct transformation.

3D Point Cloud Registration by Efficient Evolutionary Computation with Keypoint Patches Extraction

3D Point Cloud Registration by Efficient Evolutionary Computation with Keypoint Patches Extraction

基于特征匹配的三维点云配准算法

基于特征匹配的三维点云配准算法

Guaranteed Outlier Removal for Point Cloud Registration with Correspondences.

An established approach for 3D point cloud registration is to estimate the registration function from 3D keypoint correspondences. Typically, a robust technique is required to conduct the estimation, since there are false correspondences or outliers. Current 3D keypoint techniques are much less accurate than their 2D counterparts, thus they tend to produce extremely high outlier rates. A large number of putative correspondences must thus be extracted to ensure that sufficient good correspondences are available. Both factors (high outlier rates, large data sizes) however cause existing robust techniques to require very high computational cost. In this paper, we present a novel preprocessing method called guaranteed outlier removal for point cloud registration. Our method reduces the input to a smaller set, in a way that any rejected correspondence is guaranteed to not exist in the globally optimal solution. The reduction is performed using purely geometric operations which are deterministic and fast. Our method significantly reduces the population of outliers, such that further optimization can be performed quickly. Further, since only true outliers are removed, the globally optimal solution is preserved. On various synthetic and real data experiments, we demonstrate the effectiveness of our preprocessing method. Demo code is available as supplementary material, which can be found on the Computer Society Digital Library at http://doi.ieeecomputersociety.org/10.1109/TPAMI.2017.2773482.

Developing a low-cost 3D plant morphological traits characterization system

A low-cost three-dimensional (3D) plant reconstruction and morphological traits characterization system was developed. Corn plant seedlings were used as research objects for development and validation of the 3D reconstruction and point cloud data analysis algorithms. In this application, precise alignment of multiple 3D views generated by a 3D time-of-flight (ToF) sensor is critical to the 3D reconstruction of a plant. Previous research indicated that there is strong need for high-throughput, high-accuracy, and low-cost 3D plant reconstruction and trait characterization phenotyping systems. This research produced a 3D reconstruction system for indoor plant phenotyping by innovatively integrating a low-cost 2D camera, a low-cost 3D ToF camera, and a chessboard pattern beacon array to track the position and attitude of the 3D ToF sensor and, thus, accomplished precise 3D point cloud registration over multiple views. Specifically, algorithms for beacon target detection, camera pose tracking, and spatial relationship calibration between 2D and 3D cameras were developed for such a low-cost but high-performance 3D reconstruction solution. A plant analysis algorithm in a 3D space was developed to extract the morphological trait parameters of the plants by analyzing their 3D point cloud data. The phenotypical data obtained by this novel and low-cost 3D reconstruction based phenotyping system were validated by the experimental data generated by instrument and manual measurement. The results demonstrated that the developed phenotyping system has achieved promising measurement accuracy, fast processing speed while offering a low hardware cost, lending itself to a practical means of acquiring detailed 3D morphological traits for automated indoor plant phenotyping.

Efficient rotation estimation for 3D registration and global localization in structured point clouds

Fully automatic 3D point cloud registration for structured scenes is a highly challenging task. In this paper, an efficient rotation estimation algorithm is proposed for point clouds of structured scenes. This algorithm fully employs the geometric information of structured environment. For rotation estimation, a direction angle is defined for a point cloud and then the rotation matrix is obtained by comparing the difference between the distributions of angles. The proposed rotation estimation algorithm is used for both 3D registration and global localization. To conduct a full 3D registration, the translation parameters are estimated by aligning the centers of the corresponding points while the rotation parameters are estimated by the proposed algorithm. For global localization, a translation estimation algorithm is proposed using projection information. The point clouds are projected onto the orthogonal plane and template matching is performed on the projection images to calculate the translation vector. Experiments have been conducted on two datasets. Experimental results demonstrate that the proposed algorithm outperforms the state-of-the-art approaches in terms of both accuracy, computational efficiency and robustness.

An Iterative Closest Points Algorithm for Registration of 3D Laser Scanner Point Clouds with Geometric Features.

The Iterative Closest Points (ICP) algorithm is the mainstream algorithm used in the process of accurate registration of 3D point cloud data. The algorithm requires a proper initial value and the approximate registration of two point clouds to prevent the algorithm from falling into local extremes, but in the actual point cloud matching process, it is difficult to ensure compliance with this requirement. In this paper, we proposed the ICP algorithm based on point cloud features (GF-ICP). This method uses the geometrical features of the point cloud to be registered, such as curvature, surface normal and point cloud density, to search for the correspondence relationships between two point clouds and introduces the geometric features into the error function to realize the accurate registration of two point clouds. The experimental results showed that the algorithm can improve the convergence speed and the interval of convergence without setting a proper initial value.

Development of Three-Dimensional Dental Scanning Apparatus Using Structured Illumination.

We demonstrated a three-dimensional (3D) dental scanning apparatus based on structured illumination. A liquid lens was used for tuning focus and a piezomotor stage was used for the shift of structured light. A simple algorithm, which detects intensity modulation, was used to perform optical sectioning with structured illumination. We reconstructed a 3D point cloud, which represents the 3D coordinates of the digitized surface of a dental gypsum cast by piling up sectioned images. We performed 3D registration of an individual 3D point cloud, which includes alignment and merging the 3D point clouds to exhibit a 3D model of the dental cast.

3D Point Cloud Registration Based on Refined ICP Algorithm

Three-dimension (3D) digitization of human face is of great value for medical plastic surgery, video making, virtual reality, and so on. Subject to the restriction of 3D scanning equipment, a single angle of scanning measurement only obtains partial information of a face. In this paper, a multi-Kinect measurement system was proposed for the scanning of human face from multiple angles simultaneously. Firstly, different sides of human face point clouds scanned by three Kinect sensors were registered by an ICP algorithm to form a full human face point cloud. Then, a bilateral filtering algorithm was refined to remove noise in the point cloud. Finally, a point feature histogram and a refined shared nearest neighbor cluster algorithm were applied to improve the iterative speed and efficiency of the ICP algorithm. Experiment shows that the refined point cloud registration algorithm is efficient for 3D digitization of full human face.

A novel scan registration method based on the feature-less global descriptor – spherical entropy image

PurposeThis paper aims to present the spherical entropy image (SEI), a novel global descriptor for the scan registration of three-dimensional (3D) point clouds. This paper also introduces a global feature-less scan registration strategy based on SEI. It is advantageous for 3D data processing in the scenarios such as mobile robotics and reverse engineering.Design/methodology/approachThe descriptor works through representing the scan by a spherical function named SEI, whose properties allow to decompose the six-dimensional transformation into 3D rotation and 3D translation. The 3D rotation is estimated by the generalized convolution theorem based on the spherical Fourier transform of SEI. Then, the translation recovery is determined by phase only matched filtering.FindingsNo explicit features and planar segments should be contained in the input data of the method. The experimental results illustrate the parameter independence, high reliability and efficiency of the novel algorithm in registration of feature-less scans.Originality/valueA novel global descriptor (SEI) for the scan registration of 3D point clouds is presented. It inherits both descriptive power of signature-based methods and robustness of histogram-based methods. A high reliability and efficiency registration method of scans based on SEI is also demonstrated.

RESEARCH HIGHLIGHTS IN IAS

We are reviewing and commenting highlights of the research published in Image Analysis and Stereology journal (IAS), volume 35, where 16 original research papers on image analysis, computer vision, modelling, and other approaches were published. We have reported on the precision of curve length estimation in the plane. Further, a focus was on a robust estimation technique for 3D point cloud registration. Next contribution in computer vision was on the accuracy of stereo matching algorithm based on illumination control. An attempt was also made to automatically diagnose prenatal cleft lip with representative key points and identify the type of defect in three-dimensional ultrasonography. Similarly, a new report is presenting estimation of torsion of digital curves in 3D images and next, the nuchal translucency by ultrasound is being analyzed. Also in ophthalmology, image analysis may help physicians to establish a correct diagnosis, which is supported by a new approach to measure tortuosity of retinal vessel. Another report of medical significance analyzed correlation of the shape parameters for characterization of images of corneal endothelium cells. Shape analysis is also an important topic in material science, e.g. in analyzing fine aggregates in concrete. As in concrete, in fiber reinforced composites image analysis may aid in improved quality, where the direction of fibers have decisive impact on properties. Automatic defect detection using a computer vision system improves productivity quality in industrial production, hence we report of a new Haar wavelet-based approach.

An automatic 3D point cloud registration method based on regional curvature maps

3D point cloud registration is a fundamental and critical issue in 3D reconstruction and object recognition. Most of the existing methods are based on local shape descriptor. In this paper, we propose a discriminative and robust local shape descriptor-Regional Curvature Map (RCM). The keypoint and its neighboring points are firstly projected onto a 2D plane according to a robust mapping against normal errors. Then, the projection points are quantized into corresponding bins of the 2D support region and their weighted curvatures are encoded into a curvature distribution image. Based on the RCM, an efficient and accurate 3D point cloud registration method is presented. We firstly find 3D point correspondences by a RCM searching and matching strategy based on the sub-regions of the RCM. Then, a coarse registration can be achieved with geometrically consistent point correspondences, followed by a fine registration based on a modified iterative closest point (ICP) algorithm. The experimental results demonstrate that the RCM is distinctive and robust against normal errors and varying point cloud density. The corresponding registration method can achieve a higher registration precision and efficiency compared with two existing registration methods.