A Maximum Feasible Subsystem for Globally Optimal 3D Point Cloud Registration

Yu, Chanki; Ju, Da Young

doi:10.3390/s18020544

Cited by 15 publications

(10 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…First, the centroids of P S and P T are found through Equations (10) and (11), where N is the number of keypoints. Then a matrix H, similar to the covariance matrix, is accumulated by Equation (12) and decomposed to U, S, V by SVD algorithm, as described in Equation (13). The rotation matrix from source points to target points can be calculated by V and U, using Equation (14).…”

Section: Horizontal Translation Vector and Azimuth Angle Estimationmentioning

confidence: 99%

“…Registration is a fundamental and frequently encountered problem in point clouds processing [10,11]. Terrestrial LiDAR scanning scans the whole target scene station by station and registration is essential to align point clouds obtained from different stations to a unified frame [12][13][14][15][16][17][18]. In addition, the point clouds obtained by different LiDAR scanning platforms, such as TLS and MLS, are often merged to ISPRS Int.…”

Section: Introductionmentioning

confidence: 99%

“…Gao et al [33] proposed to utilize Extended Gaussian Image (EGI) to find point correspondences and thus achieve initial registration of point clouds. Yu and Ju [13] used an intrinsic shape signature keypoints detector and the Signature of Histograms of OrienTations (SHOT) [34] descriptor to find correspondences from point clouds. Then a branch-and-bound algorithm was proposed to robustly estimate registration parameters using point correspondences with incorrect matches.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Efficient Coarse Registration of Pairwise TLS Point Clouds Using Ortho Projected Feature Images

Liu

Zhang

2020

IJGI

View full text Add to dashboard Cite

The degree of automation and efficiency are among the most important factors that influence the availability of Terrestrial light detection and ranging (LiDAR) Scanning (TLS) registration algorithms. This paper proposes an Ortho Projected Feature Images (OPFI) based 4 Degrees of Freedom (DOF) coarse registration method, which is fully automated and with high efficiency, for TLS point clouds acquired using leveled or inclination compensated LiDAR scanners. The proposed 4DOF registration algorithm decomposes the parameter estimation into two parts: (1) the parameter estimation of horizontal translation vector and azimuth angle; and (2) the parameter estimation of the vertical translation vector. The parameter estimation of the horizontal translation vector and the azimuth angle is achieved by ortho projecting the TLS point clouds into feature images and registering the ortho projected feature images by Scale Invariant Feature Transform (SIFT) key points and descriptors. The vertical translation vector is estimated using the height difference of source points and target points in the overlapping regions after horizontally aligned. Three real TLS datasets captured by the Riegl VZ-400 and the Trimble SX10 and one simulated dataset were used to validate the proposed method. The proposed method was compared with four state-of-the-art 4DOF registration methods. The experimental results showed that: (1) the accuracy of the proposed coarse registration method ranges from 0.02 m to 0.07 m in horizontal and 0.01 m to 0.02 m in elevation, which is at centimeter-level and sufficient for fine registration; and (2) as many as 120 million points can be registered in less than 50 s, which is much faster than the compared methods.

show abstract

Section: Horizontal Translation Vector and Azimuth Angle Estimationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Efficient Coarse Registration of Pairwise TLS Point Clouds Using Ortho Projected Feature Images

Liu

Zhang

2020

IJGI

View full text Add to dashboard Cite

show abstract

“…In recent years, there has been a series of work utilizing BnB to globally estimate the rigid transformation between two point sets, such as Go-ICP [ 24 ], GOGMA [ 25 ] and so on [ 34 , 35 , 36 ]. However, these methods focus on rigid point set registration and they cannot be simply extended to solve the 7-DoF similarity registration problem.…”

Section: Related Workmentioning

confidence: 99%

Efficient Similarity Point Set Registration by Transformation Decomposition

Wang

Chen

Wang

2020

Sensors

View full text Add to dashboard Cite

Point set registration is one of the basic problems in computer vision. When the overlap ratio between point sets is small or the relative transformation is large, local methods cannot guarantee the accuracy. However, the time complexity of the branch and bound (BnB) optimization used in most existing global methods is exponential in the dimensionality of parameter space. Therefore, seven-Degrees of Freedom (7-DoF) similarity transformation is a big challenge for BnB. In this paper, a novel rotation and scale invariant feature is introduced to decouple the optimization of translation, rotation, and scale in similarity point set registration, so that BnB optimization can be done in two lower dimensional spaces. With the transformation decomposition, the translation is first estimated and then the rotation is optimized by maximizing a robust objective function defined on consensus set. Finally, the scale is estimated according to the potential correspondences in the obtained consensus set. Experiments on synthetic data and clinical data show that our method is approximately two orders of magnitude faster than the state-of-the-art global method and more accurate than a typical local method. When the outlier ratio with respect to the inliers is up to 1.0, our method still achieves accurate registration.

show abstract

“…A maxFS problem is at the heart of the registration problem of partially overlapped 3 D surfaces in computer vision, computer graphics, and robotics (Yu and Ju 2018). When the point clouds registered by sensors have noise or outliers it is difficult to make pairwise registrations for the points.…”

Section: Other Recent Applicationsmentioning

confidence: 99%

The maximum feasible subset problem (maxFS) and applications

Chinneck

2019

INFOR: Information Systems and Operational Research

View full text Add to dashboard Cite

The maximum feasible subset problem (maxFS) is this: given an infeasible set of constraints, find a largest cardinality subset that admits a feasible solution. This problem is NP-hard but has been studied extensively for the case of linear constraints, and good heuristic solution algorithms are available. There is a surprisingly large range of applications for algorithms that solve the linear maxFS problem, including analyzing infeasible linear programs, finding the data depth, placing separating hyperplanes in classification decision trees, recovering sparse data in compressed sensing, dimension reduction in nonnegative matrix factorization, etc. This paper reviews maxFS solution algorithms, and surveys the existing and new applications.

show abstract

A Maximum Feasible Subsystem for Globally Optimal 3D Point Cloud Registration

Cited by 15 publications

References 37 publications

Efficient Coarse Registration of Pairwise TLS Point Clouds Using Ortho Projected Feature Images

Efficient Coarse Registration of Pairwise TLS Point Clouds Using Ortho Projected Feature Images

Efficient Similarity Point Set Registration by Transformation Decomposition

The maximum feasible subset problem (maxFS) and applications

Contact Info

Product

Resources

About