LFA-Net: Enhanced PointNet and Assignable Weights Transformer Network for Partial-to-Partial Point Cloud Registration

Terrestrial laser scanners have achieved great popularity in the last decade. Their easy on-site application and the possibility of flexible and high quality post processing added to their success in several fields such as architectural, archaeological, and heritage documentation. We present a method for handling the automatic registration of point clouds which are characterized by significant noise level, generally imperfect geometry and occlusions. Hereby we combine and extend already existing and established methods to facilitate the registration of point clouds without prior pre-processing. Our approach consists—similar to other methods—of three steps which are scan analysis, pair-wise registration, and global registration. To handle the abovementioned datasets we propose to use imperfect and subdivided features, and to implement Genetic Algorithms (GAs). At the same time our approach can be seen as extension to already known Genetic Algorithms used for the registration of point clouds. By implementing an adapted version of a Genetic Algorithm in the classical registration process between rough alignment and fine registration, we are able to maintain robustness and computational performance also when registering point clouds of bigger objects characterized by a notably increased number of points, a significant noise level, and occlusions. We show and discuss the successful application of the algorithm on a scene which does not consist of classical geometric primitives such as planes.

Abstract. The registration of ground laser point cloud and close-range image is the key content of high-precision 3D reconstruction of cultural relic object. In view of the requirement of high texture resolution in the field of cultural relic at present, The registration of point cloud and image data in object reconstruction will result in the problem of point cloud to multiple images. In the current commercial software, the two pairs of registration of the two kinds of data are realized by manually dividing point cloud data, manual matching point cloud and image data, manually selecting a two – dimensional point of the same name of the image and the point cloud, and the process not only greatly reduces the working efficiency, but also affects the precision of the registration of the two, and causes the problem of the color point cloud texture joint. In order to solve the above problems, this paper takes the whole object image as the intermediate data, and uses the matching technology to realize the automatic one-to-one correspondence between the point cloud and multiple images. The matching of point cloud center projection reflection intensity image and optical image is applied to realize the automatic matching of the same name feature points, and the Rodrigo matrix spatial similarity transformation model and weight selection iteration are used to realize the automatic registration of the two kinds of data with high accuracy. This method is expected to serve for the high precision and high efficiency automatic 3D reconstruction of cultural relic objects, which has certain scientific research value and practical significance.

We consider the problem of online and real-time registration of partial point clouds obtained from an unseen real-world rigid object without knowing its 3D model. The point cloud is partial as it is obtained by a depth sensor capturing only the visible part of the object from a certain viewpoint. It introduces two main challenges: 1) two partial point clouds do not fully overlap and 2) keypoints tend to be less reliable when the visible part of the object does not have salient local structures. To address these issues, we propose DeepPRO, a keypoint-free and an end-to-end trainable deep neural network. Its core idea is inspired by how humans align two point clouds: we can imagine how two point clouds will look like after the registration based on their shape. To realize the idea, DeepPRO has inputs of two partial point clouds and directly predicts the point-wise location of the aligned point cloud. By preserving the ordering of points during the prediction, we enjoy dense correspondences between input and predicted point clouds when inferring rigid transform parameters. We conduct extensive experiments on the real-world Linemod and synthetic ModelNet40 datasets. In addition, we collect and evaluate on the PRO1k dataset, a large-scale version of Linemod meant to test generalization to real-world scans. Results show that DeepPRO achieves the best accuracy against thirteen strong baseline methods, e.g., 2.2mm ADD on the Linemod dataset, while running 50 fps on mobile devices.

During the process of the point cloud registration, the problem that the accuracy is not high is due to the unknown relative position of the multi-view point clouds and the diversity of the target structure. This paper proposes a point cloud registration method based on extracting overlapping regions to solve it. First of all, according to the characteristics of the target geometric structure, each point cloud is divided into blocks and the ESF multi-dimensional shape descriptors of each point cloud block are constructed, the region with the greatest similarity between the descriptors, namely the overlapping region between the point clouds. Then our algorithm uses the Super Four Point Fast Robust Matching (Super4PCS) algorithm to execute a coarse registration of point clouds in overlapping regions, the initial positions of the point clouds after the coarse registration are obtained according to the consistency constraints, and finally uses the iterative closest point algorithm (ICP) to precisely register the overlapping region and obtain the final point cloud model. Compared with traditional Super4PCS algorithm, the experimental results show that the method proposed in this paper effectively improves the accuracy and accelerates the process of point cloud registration.

The challenges of point cloud registration in intelligent vehicle driving lie in the large scale, complex distribution, high noise, and strong sparsity of lidar point cloud data. This paper proposes an efficient registration algorithm for large-scale outdoor road scenes by selecting the continuous distribution of key area laser point clouds as the registration point cloud. The algorithm extracts feature descriptions of the key point cloud and introduces local geometric features of the point cloud to complete rough and fine registration under constraints of key point clouds and point cloud features. The algorithm is verified through extensive experiments under multiple scenarios, with an average registration time of 0.5831 s and an average accuracy of 0.06996 m, showing significant improvement compared to other algorithms. The algorithm is also validated through real-vehicle experiments, demonstrating strong versatility, reliability, and efficiency. This research has the potential to improve environment perception capabilities of autonomous vehicles by solving the point cloud registration problem in large outdoor scenes.

Point cloud registration, which estimates a rigid transformation matrix between two point clouds, is a fundamental process in numerous applications. While existing detector-free techniques present exceptional performance, they overlook the extraction of hybrid local features that capture correlations between points and their neighbours, thereby limiting the quality of point cloud recognition. Moreover, these approaches typically treat point clouds as sequential data and employ the transformer to integrate global context from all points, which inevitably introduces interference from irrelevant regions, hence affecting the registration accuracy. In this work, we propose a novel detector-free approach AGHL to address these challenges. For the first issue, AGHL introduces a hybrid local feature perception module that designs two parallel branches to concurrently extract low-level and high-level local features, which effectively encode the correlations between each point and its neighborhood points in both Euclidean space and high-dimensional feature space. For the second issue, AGHL develops an anchor-guided cross attention that adheres to the local geometric consistency to constrain the network's attention on reliable anchors, thereby effectively suppressing interference from irrelevant regions. Benefiting from these techniques, AGHL achieves impressive point cloud registration accuracy across all synthetic, indoor, and outdoor datasets. Furthermore, we build an experimental platform and conduct a real-world robot localization experiment, with results showing the strong generalization ability of AGHL.

Coded light measurement is an important method of non-contact optical measurement. Because of the camera view limitations, it was hard to complete the measurement of the whole surface at a time. Then it needs registration of 3D point clouds. The registration of point clouds can reflect the 3D surface of the object to be measured, which can achieve the full range of requirements. According to the encoded optical 3D measurement system has been built in laboratory, exploring a method which is high accurate and high speed to reach a full range measurement to industrial parts. This paper describes the color target cube. According to the existing method of coordinate transformation (the Nearest Point Iterative Method), Evaluating the registration accuracy. The registration method based on auxiliary three-dimensional targets was been validated, which has its own rationality.

With the widespread application of large-scale 3D point cloud data in real-world scenarios, efficient and accurate point cloud registration has become a crucial challenge. We propose an end-to-end point cloud registration method based on the Transformer architecture. This method addresses the issues of low overlap and registration in large scenes, exhibiting strong algorithmic versatility and efficiency. We introduce a combination of dynamic position encoding and ternary angular position encoding within the Transformer, effectively enhancing the representation capability of point cloud data and algorithmic generality, thus better tackling point cloud registration challenges in large scenes. Additionally, to enhance the learning capacity of the attention mechanism, we employ an improved cross-attention mechanism that multiplies the softmax with adaptive weights, enabling the model to capture key information within the point cloud more accurately. In the decoding stage, we introduce a multi-scale feature fusion approach that fully exploits the multi-layer information in point cloud data, further improving registration accuracy and robustness. Through the fusion of multi-scale features, we effectively mitigate information loss and handle matching problems between point clouds of varying sizes. Experimental results demonstrate the excellence of our method in addressing low overlap and registration tasks in large scenes, validated across multiple datasets including 3DMatch, ModelNet, KITTI, and MVP-RG.

The present work proposes a new method Reg3DFacePtCd for registration of point clouds. The key contribution of the present method is that an unknown face in 3D point cloud form is given to the system and is registered to the already existing known 3D face point clouds using a fast 3D face registration method. The novelty of the present technique is that at first the alignment and registration parameters are found out by initially registering eight key points of the unknown source model to that of the known model. Next, the rest of the point clouds of the unknown model are registered to that of the known model using the same parameters found as above. The main method used for alignment is iterative closest point (ICP) using point-based technique followed by registration in the least squares sense. Mainly there are two significant contributions. Firstly, we have developed a new mathematical model facial landmark point based model across poses to obtain the target or the known model to which all the unknown models will be registered. Secondly, a novel way to accelerate point cloud matching by reducing the number of points has been proposed. Using a small number of points necessarily would speed up the registration process but may inculcate errors. So, to determine the registration quality of the fundamental eight key points on which the entire registration process is based, a new robust metric namely ICV (ICP certainty vector) consisting of several key components have been used. Finally, we have addressed several important face registration issues like pre-processing, convergence and quality of registration of the entire facial point cloud based on the eight key points. Extensive experimentation on Frav3D, GavabDB, and the Bosphorus databases on a high-performance computing environment show the novelty and robustness of the method.

Point cloud registration is of great significance to the reconstruction of high‐precision 3D city models. There are some challenges when aligning aerial and close‐range photogrammetric point clouds, such as huge view differences caused by the different sights of the sensors, massive noisy points due to the error in dense matching, and scale uncertainty since there are no control points for absolute orientation. To achieve complementary advantages of aerial and close‐range point clouds, in this paper, a robust cross‐source point clouds registration method is proposed using image visual context features and scale consistency. First, a cross‐view image matching method based on image visual context features is proposed to obtain corresponding points. Second, to overcome the challenges of noisy points and scale differences, an outlier filtering method is designed based on scale consistency. Finally, the dual quaternions model considering the scale factor is introduced to solve the spatial transformation model. To analyze the feasibility of this method qualitatively and quantitatively, experiments are conducted using the public scene dataset of Dortmund, Germany, and the scene dataset of Zhengzhou City, China. Three kinds of cross‐source point clouds registration experiments are conducted in this paper, including aerial and close‐range point clouds registration in Dortmund, and aerial and ground point clouds registrations in both Dortmund and Zhengzhou. The chamfer distances of the three sets of experiments are 4.48 m, 5.97 m and 4.78 m, respectively. The ablation study shows that the outlier filtering method and the dual quaternions model improve the accuracy by at least 14% and 25%, respectively. The experiments demonstrate that the method accomplishes the cross‐source point clouds registration in large‐scale scenes accurately and efficiently, providing a solid foundation for subsequent fine 3D reconstruction.

Stockpile volume estimation plays a critical role in several industrial/commercial bulk material management applications. LiDAR systems are commonly used for this task. Thanks to Global Navigation Satellite System (GNSS) signal availability in outdoor environments, Uncrewed Aerial Vehicles (UAV) equipped with LiDAR are frequently adopted for the derivation of dense point clouds, which can be used for stockpile volume estimation. For indoor facilities, static LiDAR scanners are usually used for the acquisition of point clouds from multiple locations. Acquired point clouds are then registered to a common reference frame. Registration of such point clouds can be established through the deployment of registration targets, which is not practical for scalable implementation. For scans in facilities bounded by planar walls/roofs, features can be automatically extracted/matched and used for the registration process. However, monitoring stockpiles stored in dome facilities remains to be a challenging task. This study introduces an image-aided fine registration strategy of acquired sparse point clouds in dome facilities, where roof and roof stringers are extracted, matched, and modeled as quadratic surfaces and curves. These features are then used in a Least Squares Adjustment (LSA) procedure to derive well-aligned LiDAR point clouds. Planar features, if available, can also be used in the registration process. Registered point clouds can then be used for accurate volume estimation of stockpiles. The proposed approach is evaluated using datasets acquired by a recently developed camera-assisted LiDAR mapping platform—Stockpile Monitoring and Reporting Technology (SMART). Experimental results from three datasets indicate the capability of the proposed approach in producing well-aligned point clouds acquired inside dome facilities, with a feature fitting error in the 0.03–0.08 m range.

The robustness of correspondence-based point cloud registration relies on transformation invariance and intrinsic distinctiveness of the descriptors computed for registration. However, for challenging scenarios with different objects having similar local geometry and low point cloud overlap, existing descriptors struggle to retain these properties. This inevitably leads to a considerable loss in matching inlier rates and compromised registration. To address the issue, we propose RID-Net that computes robust Rotation-Invariant and Distinctive descriptors for point cloud registration. Our model works on the philosophy of locally accurate feature abstraction for the points while also accounting for the surroundings. Specifically, the locally accurate representation generated by our rotation-invariant MLP-based module is further enhanced through a proposed spherical shell attention mechanism to explicitly focus on the surrounding geometry. The overall network is designed by staking the proposed component, while eventually leveraging coarse-to-fine correspondences and a robust pose estimator to compute the ultimate output. With extensive evaluation on indoor, outdoor and synthetic point cloud data, we establish the efficacy of the proposed RID-Net. In particular, our method improves the inlier ratio by about 4% on 3DMatch/3DLoMatch dataset and reduces the rotation and translation errors by 26.1% and 19.7% on the KITTI dataset.

The registration of 3-D point clouds is an important procedure during the terrestrial laser scanning data processing. Recently, due to their high flexibility and the powerful mathematical model, a large amount of least-squares-based (LSs-based) methods are proposed and widely applied to estimate the transformation parameters of 3-D point clouds registration. In these LSs-based methods some based on the generalized Gauss–Markov model do not correct the influence of random errors on source 3-D point clouds. Although there are other methods based on the errors-in-variables (EIV) model, they are inapplicable for transformation problems with large rotation angles and arbitrary scale ratio. In addition, the gross errors are usually ignored in previous studies on 3-D point clouds registration, which, however, exists commonly and could distort the registration severely. Aiming to avoid the influence of gross errors and extend its application, an advanced outlier detected total least-squares (OD-TLS) method is proposed in this paper. Based on the generalized EIV model OD-TLS performs a seven-parameter 3-D similarity transformation with large rotation angles and arbitrary scale ratio. The random errors of both source and target 3-D point clouds are considered. Furthermore, outliers are detected and removed automatically by combining the data snooping method with total least-squares (TLS) estimation. In order to indicate the benefits of OD-TLS, comparative experiments with the LS3D and weighted total least squares (WTLS) on synthetic and real-world scanned 3-D point clouds were performed. The experimental results show OD-TLS not only enhances the registration accuracy but also increases its robustness.

Laser three-dimensional (3D) image is a novel non-cooperative target 3D image acquisition technology, and the improvements in detection capability and imaging accuracy of the system are critically dependent on efficient echo-signal processing technique and 3D reconstruction method. The registration process is an essential step in array 3D imaging laser point cloud data processing. Registration of point clouds is an effective method that solves the problem caused by the target self-occlusion in the laser 3D imaging system. The accurate registration result will help provide better support for subsequent applications, such as object reconstruction and target recognition. In this study, a set of thresholds in the iterative closest point (ICP) algorithm is analysed on the basis of the characteristics of the laser array 3D imaging system and is combined with the range error and visual lateral resolution of the system, which are both important parameters in the imaging system. To improve the accuracy and speed of registration, the stop threshold of the iterative algorithm and the corresponding point-distance threshold in the algorithm are established in a novel way based on the range error and visual lateral resolution of the system. This forms the foundation, based on which an adaptive threshold ICP algorithm is proposed. The principal idea of the algorithm is to improve the threshold set that has a considerable effect on the accuracy and speed of registration. At first, the characteristics of the imaging point clouds of the laser array 3D imaging system are analysed in the algorithm. Based on this analysis, the distance between the two point clouds and corresponding points with ideal registrations are estimated theoretically, according to the range error and visual lateral resolution of the system. The simulation results show that the theoretically estimated results and actual results have the same variation tendency, thus providing a theoretical basis for subsequent improvements. Next, the estimated results are added according to the iterative closest point algorithm. This implies that the registration thresholds are capable of changing and adapting under different iterations and imaging systems, thus improving the speed and accuracy of registrations. This phenomenon is not seen in other algorithms. Experiments involving laser array imaging of a point cloud and laser scanning of depth imaging data show that the algorithm is practical and effective for both imaging types of point clouds and can improve the speed and accuracy of registration notably. The effectiveness and feasibility of the proposed algorithm are thus verified. In addition, for its full consideration of the imaging system, the basic idea of the proposed algorithm can be used for designing future applications as required.

Scanning electron microscopy (SEM) has an important application in the petroleum field, which is often used to analyze the microstructure of reservoir rocks, etc. Most of these analyses are based on two-dimensional images. In fact, SEM can carry out micro-nano scale three-dimensional measurement, and three-dimensional models can provide more accurate information than two-dimensional images. Among the commonly used SEM 3D reconstruction methods, parallax depth mapping is the most commonly used method. Multiple SEM images can be obtained by continuously tilting the sample table at a certain Angle, and multiple point clouds can be generated according to the parallax depth mapping method, and a more complete point clouds recovery can be achieved by combining the point clouds registration. However, the root mean square error of the point clouds generated by this method is relatively large and unstable after participating in point clouds registration. Therefore, this paper proposes a new method for generating point clouds. Firstly, the sample stage is rotated by a certain angle to obtain two SEM images. This operation makes the rotation matrix a known quantity. Then, based on the imaging model, an equation system is constructed to estimate the unknown translation parameters, and finally, triangulation is used to obtain the point clouds. The method proposed in this paper was tested on a publicly available 3D SEM image set, and the results showed that compared to the disparity depth mapping method, the point clouds generated by our method showed a significant reduction in root mean square error and relative rotation error in point clouds registration.