A Point Cloud Registration Algorithm Based on Feature Extraction and Matching

Liu, Yongshan; Kong, Dehan; Zhao, Dandan; Gong, Xiang; Han, Guichun

doi:10.1155/2018/7352691

Cited by 23 publications

(19 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…erefore, many modifications related to the NDT algorithm were proposed by researchers. For example, Yong et al [201] introduced an improved normal distribution transform algorithm in a precise registration step. In this method, PDF is replaced by a mixed probability density function.…”

Section: Normal Distributionmentioning

confidence: 99%

A Tutorial Review on Point Cloud Registrations: Principle, Classification, Comparison, and Technology Challenges

Wang

Zhang

2021

Mathematical Problems in Engineering

View full text Add to dashboard Cite

A point cloud as a collection of points is poised to bring about a revolution in acquiring and generating three-dimensional (3D) surface information of an object in 3D reconstruction, industrial inspection, and robotic manipulation. In this revolution, the most challenging but imperative process is point could registration, i.e., obtaining a spatial transformation that aligns and matches two point clouds acquired in two different coordinates. In this survey paper, we present the overview and basic principles, give systematical classification and comparison of various methods, and address existing technical problems in point cloud registration. This review attempts to serve as a tutorial to academic researchers and engineers outside this field and to promote discussion of a unified vision of point cloud registration. The goal is to help readers quickly get into the problems of their interests related to point could registration and to provide them with insights and guidance in finding out appropriate strategies and solutions.

show abstract

Section: Normal Distributionmentioning

confidence: 99%

A Tutorial Review on Point Cloud Registrations: Principle, Classification, Comparison, and Technology Challenges

Wang

Zhang

2021

Mathematical Problems in Engineering

View full text Add to dashboard Cite

show abstract

“…However, in the presented work, it is suggested that the objective function incorporates ego-motion estimation with uncertainty, thus its performance is revealed to be more accurate under feature-poor and self-similar environments. In 2018, Liu et al [ 22 ] proposed an Improved NDT (INDT) that registers using fractional pre-processed feature points only. For this, the work applied a Fast Point Feature Histogram (FPFH) descriptor and Hausdorff distance method to extract feature points.…”

Section: Related Workmentioning

confidence: 99%

“…The adapted score function of the proposed algorithm has similarities with the previous reference [ 21 ], yet the objective function employs a unique classifier coefficient depending on point characteristics. The INDT approach in [ 22 ] also extracts and uses feature points during the scan matching procedure, yet differs from the proposed methods in that extracted feature points are only used for the registration process. Due to this reasoning, the INDT in [ 22 ] is excluded from the performance comparison algorithm because poor performance is achieved through INDT with a relatively smaller number of feature points for the 2D measurement environment.…”

Section: Related Workmentioning

confidence: 99%

“…The INDT approach in [ 22 ] also extracts and uses feature points during the scan matching procedure, yet differs from the proposed methods in that extracted feature points are only used for the registration process. Due to this reasoning, the INDT in [ 22 ] is excluded from the performance comparison algorithm because poor performance is achieved through INDT with a relatively smaller number of feature points for the 2D measurement environment. On the other hand, the NDT-ICP method in [ 23 ], which performs NDT and ICP sequentially, is implemented as a performance comparison algorithm.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Integrated Pose Estimation Using 2D Lidar and INS Based on Hybrid Scan Matching

Park

Lee

Sung

2021

Sensors

View full text Add to dashboard Cite

Point cloud data is essential measurement information that has facilitated an extended functionality horizon for urban mobility. While 3D lidar and image-depth sensors are superior in implementing mapping and localization, sense and avoidance, and cognitive exploration in an unknown area, applying 2D lidar is inevitable for systems with limited resources of weight and computational power, for instance, in an aerial mobility system. In this paper, we propose a new pose estimation scheme that reflects the characteristics of extracted feature point information from 2D lidar on the NDT framework for exploiting an improved point cloud registration. In the case of the 2D lidar point cloud, vertices and corners can be viewed as representative feature points. Based on this feature point information, a point-to-point relationship is functionalized and reflected on a voxelized map matching process to deploy more efficient and promising matching performance. In order to present the navigation performance of the mobile object to which the proposed algorithm is applied, the matching result is combined with the inertial navigation through an integration filter. Then, the proposed algorithm was verified through a simulation study using a high-fidelity flight simulator and an indoor experiment. For performance validation, both results were compared and analyzed with the previous techniques. In conclusion, it was demonstrated that improved accuracy and computational efficiency could be achieved through the proposed algorithms.

show abstract

“…Experiments show that this method achieved better robustness in nonrigid registration. In recent years, under the framework of EM algorithm optimization, some robust point set registration algorithms have been proposed [10,12,13].…”

Section: Introductionmentioning

confidence: 99%

Point Set Registration Based on Improved KL Divergence

Lee

2021

Scientific Programming

View full text Add to dashboard Cite

A point set registration algorithm based on improved Kullback–Leibler (KL) divergence is proposed. Each point in the point set is represented as a Gaussian distribution. The Gaussian distribution contains the position information of the candidate point and surrounding ones. In this way, the entire point set can be modeled as a Gaussian mixture model (GMM). The registration problem of two point sets is further converted as a minimization problem of the improved KL divergence between two GMMs, and the genetic algorithm is used to optimize the solution. Experimental results show that the proposed algorithm has strong robustness to noise, outliers, and missing points, which achieves better registration accuracy than some state-of-the-art methods.

show abstract

A Point Cloud Registration Algorithm Based on Feature Extraction and Matching

Cited by 23 publications

References 11 publications

A Tutorial Review on Point Cloud Registrations: Principle, Classification, Comparison, and Technology Challenges

A Tutorial Review on Point Cloud Registrations: Principle, Classification, Comparison, and Technology Challenges

Integrated Pose Estimation Using 2D Lidar and INS Based on Hybrid Scan Matching

Point Set Registration Based on Improved KL Divergence

Contact Info

Product

Resources

About