2016
DOI: 10.1016/j.mechatronics.2015.10.014
|View full text |Cite
|
Sign up to set email alerts
|

3D reconstruction and multiple point cloud registration using a low precision RGB-D sensor

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1

Citation Types

0
22
0

Year Published

2016
2016
2023
2023

Publication Types

Select...
4
4
2

Relationship

0
10

Authors

Journals

citations
Cited by 45 publications
(23 citation statements)
references
References 11 publications
0
22
0
Order By: Relevance
“…The tested sensor was found to be a low-cost solution capable of monitoring multi-component displacement with an error of less than 5% for displacements larger than 10 mm. Optical methods have also been used by Henry et al (2014) for 3D modeling of indoor environments, and by (Takimoto et al, 2016) for 3D reconstruction of various objects. J auregui et al (2003) reported that photogrammetry performed with semi-metric digital cameras has accuracy ranging from 2-10% of the measured vertical displacements.…”
Section: Introductionmentioning
confidence: 99%
“…The tested sensor was found to be a low-cost solution capable of monitoring multi-component displacement with an error of less than 5% for displacements larger than 10 mm. Optical methods have also been used by Henry et al (2014) for 3D modeling of indoor environments, and by (Takimoto et al, 2016) for 3D reconstruction of various objects. J auregui et al (2003) reported that photogrammetry performed with semi-metric digital cameras has accuracy ranging from 2-10% of the measured vertical displacements.…”
Section: Introductionmentioning
confidence: 99%
“…Point feature based: In the application of rigid transformation estimation, the most common technique is the random sample consensus (RANSAC) algorithm [19,20] which randomly selects three pairs of matched feature points to compute a 6DoF transformation T = (R, t), then applies the solved T to other correspondences and count the number of candidates which follow transformation T. Repeats this process for N times and chooses the candidate set contains maximum inlier count and solves a transformation T* from the set as the estimated initial alignment for iterative closest point (ICP) algorithm [21,22]. This RANSAC +ICP paradigm is the most popular solution to rigid registration problem which has found wide applications [7,8,14]. Besides, Papazov and Burschka [23] combine a robust descriptor which originates from the distance and angle-preserving properties of rigid transformation, a hash table used to retrieve corresponding pairs in model and scene and efficient RANSAC variant to generate and verify transformation hypothesis.…”
Section: Related Workmentioning
confidence: 99%
“…Among them, the ICP algorithm is widely used for rigid registration due to its simplicity, accuracy and speed. Moreover, with the development of imaging technology, some scholars focus on the registration of RGB-D (red, green, blue and depth) data [15]- [17]. In this paper, our work mainly focuses on the problem of registering the RGB-D data based on ICP registration framework.…”
Section: Introductionmentioning
confidence: 99%