SCVP: Learning One-shot View Planning via Set Covering for Unknown Object Reconstruction

2022 ◽  
pp. 1-1
Author(s):  
Sicong Pan ◽  
Hao Hu ◽  
Hui Wei
Keyword(s):  
Set Covering ◽  
Object Reconstruction ◽  
View Planning ◽  
Unknown Object

scholarly journals Model-based view planning for building inspection and surveillance using voxel dilation, Medial Objects, and Random-Key Genetic Algorithm

2017 ◽  
Vol 5 (3) ◽  
pp. 337-347 ◽  
Author(s):  
Wei Jing ◽  
Kenji Shimada
Keyword(s):  
Genetic Algorithm ◽  
Computational Method ◽  
Set Covering ◽  
Covering Problem ◽  
Set Covering Problem ◽  
Greedy Search ◽  
View Planning ◽  
3D Environment ◽  
Optimal Set ◽  
Gaussian Sampling

Abstract Model-based view planning is to find a near-optimal set of viewpoints that cover the surface of a target geometric model. It has been applied to many building inspection and surveillance applications with Unmanned Aerial Vehicle (UAV). Previous approaches proposed in the past few decades suffer from several limitations: many of them work exclusively for 2D problems, generate only a sub-optimal set of views for target surfaces in 3D environment, and/or generate a set of views that cover only part of the target surfaces in 3D environment. This paper presents a novel two-step computational method for finding near-optimal views to cover the surface of a target set of buildings using voxel dilation, Medial Objects (MO), and Random-Key Genetic Algorithm (RKGA). In the first step, the proposed method inflates the building surfaces by voxel dilation to define a sub-volume around the buildings. The MO of this sub-volume is then calculated, and candidate viewpoints are sampled using Gaussian sampling around the MO surface. In the second step, an optimization problem is formulated as (partial) Set Covering Problem and solved by searching through the candidate viewpoints using RKGA and greedy search. The performance of the proposed two-step computational method was measured with several computational cases, and the performance was compared with two previously proposed methods: the optimal-scan-zone method and the randomized sampling-based method. The results demonstrate that the proposed method outperforms the previous methods by finding a better solution with fewer viewpoints and higher coverage ratio compared to the previous methods. Highlights A two-step “generate-test” view planning method is proposed. Voxel dilation, Medial Objects and Gaussian sampling are used to generate viewpoints. Random-Key GA and Greedy search are combined to solve the Set Covering Problem. The proposed method is benchmarked and outperforms two existing methods.

Volumetric view planning for 3D reconstruction with multiple manipulators

2015 ◽  
Vol 42 (6) ◽  
pp. 533-543 ◽  
Author(s):  
Liangzhi Li ◽  
Nanfeng Xiao
Keyword(s):  
Three Dimensional ◽  
Industrial Applications ◽  
Computationally Efficient ◽  
Object Reconstruction ◽  
View Planning ◽  
Reconstruction Process ◽  
Content Type ◽  
Speed Up ◽  
Next Best View ◽  
Multiple Manipulators

Purpose – This paper aims to propose a new view planning method which can be used to calculate the next-best-view (NBV) for multiple manipulators simultaneously and build an automated three-dimensional (3D) object reconstruction system, which is based on the proposed method and can adapt to various industrial applications. Design/methodology/approach – The entire 3D space is encoded with octree, which marks the voxels with different tags. A set of candidate viewpoints is generated, filtered and evaluated. The viewpoint with the highest score is selected as the NBV. Findings – The proposed method is able to make the multiple manipulators, equipped with “eye-in-hand” RGB-D sensors, work together to accelerate the object reconstruction process. Originality/value – Compared to the existed approaches, the proposed method in this paper is fast, computationally efficient, has low memory cost and can be used in actual industrial productions where the multiple different manipulators exist. And, more notably, a new algorithm is designed to speed up the generation and filtration of the candidate viewpoints, which can guarantee both speed and quality.

View planning for automated three-dimensional object reconstruction and inspection

2003 ◽  
Vol 35 (1) ◽  
pp. 64-96 ◽  
Author(s):  
William R. Scott ◽  
Gerhard Roth ◽  
Jean-François Rivest
Keyword(s):  
Three Dimensional ◽  
Object Reconstruction ◽  
View Planning ◽  
Dimensional Object

scholarly journals Volumetric Next-best-view Planning for 3D Object Reconstruction with Positioning Error

10.5772/58759 ◽  
2014 ◽  
Vol 11 (10) ◽  
pp. 159 ◽  
Author(s):  
J. Irving Vasquez-Gomez ◽  
L. Enrique Sucar ◽  
Rafael Murrieta-Cid ◽  
Efrain Lopez-Damian
Keyword(s):  
Positioning Error ◽  
Object Reconstruction ◽  
View Planning ◽  
3D Object ◽  
3D Object Reconstruction ◽  
Next Best View

scholarly journals An active reconstruction algorithm based on partial prior information

2020 ◽  
Vol 17 (1) ◽  
pp. 172988142090420
Author(s):  
Yanzi Kong ◽  
Feng Zhu ◽  
Yingming Hao ◽  
Haibo Sun ◽  
Yilin Xie ◽  
...  
Keyword(s):  
Prior Information ◽  
Active Vision ◽  
Reconstruction Algorithm ◽  
Shape Estimation ◽  
Object Reconstruction ◽  
View Planning ◽  
Visual Tasks ◽  
Motion Paths ◽  
Active Reconstruction ◽  
Next Best View

Active reconstruction is an intelligent perception method that achieves object modeling with few views and short motion paths by systematically adjusting the parameters of the camera while ensuring model integrity. Part of the object information is always known for vision tasks in real scenes, and it provides some guidance for the view planning. A two-step active reconstruction algorithm based on partial prior information is presented, which includes rough shape estimation phase and complete object reconstruction phase, and both of them introduce the concept of active vision. An information expression method is proposed that can be used to manually initialize the repository according to specific visual tasks, and then the prior information and detected information are used to plan the next best view online until the object reconstruction is completed. The method is evaluated with simulated experiments and the result is compared with other methods.

scholarly journals Active Object Reconstruction Using a Guided View Planner

2018 ◽  
Author(s):  
Xin Yang ◽  
Yuanbo Wang ◽  
Yaru Wang ◽  
Baocai Yin ◽  
Qiang Zhang ◽  
...  
Keyword(s):  
Information Acquisition ◽  
3D Model ◽  
Alternative Methods ◽  
Active Object ◽  
Reconstruction Accuracy ◽  
Object Reconstruction ◽  
View Planning ◽  
Aggregate Information ◽  
Active Reconstruction ◽  
Reconstruction Model

Inspired by the recent advance of image-based object reconstruction using deep learning, we present an active reconstruction model using a guided view planner. We aim to reconstruct a 3D model using images observed from a planned sequence of informative and discriminative views. But where are such informative and discriminative views around an object? To address this we propose a unified model for view planning and object reconstruction, which is utilized to learn a guided information acquisition model and to aggregate information from a sequence of images for reconstruction. Experiments show that our model (1) increases our reconstruction accuracy with an increasing number of views (2) and generally predicts a more informative sequence of views for object reconstruction compared to other alternative methods.

Sign in / Sign up

Export Citation Format

Share Document