2021 IEEE International Conference on Robotics and Automation (ICRA) 2021
DOI: 10.1109/icra48506.2021.9561855
|View full text |Cite
|
Sign up to set email alerts
|

City-scale Scene Change Detection using Point Clouds

Abstract: We propose a method for detecting structural changes in a city using images captured from vehicular mounted cameras over traversals at two different times. We first generate 3D point clouds for each traversal from the images and approximate GNSS/INS readings using Structure-from-Motion (SfM). A direct comparison of the two point clouds for change detection is not ideal due to inaccurate geo-location information and possible drifts in the SfM. To circumvent this problem, we propose a deep learning-based non-rig… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1

Citation Types

0
3
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
6
1

Relationship

0
7

Authors

Journals

citations
Cited by 11 publications
(3 citation statements)
references
References 40 publications
0
3
0
Order By: Relevance
“…More recently, the state-of-the-art includes methods focusing on the comparison of three-dimensional point cloud pairs, taken at different times (Yew and Lee, 2021). In the latter paper the authors infer, using a neural network, a non-rigid transform to temporally align and account for point cloud inconsistencies.…”
Section: Related Workmentioning
confidence: 99%
“…More recently, the state-of-the-art includes methods focusing on the comparison of three-dimensional point cloud pairs, taken at different times (Yew and Lee, 2021). In the latter paper the authors infer, using a neural network, a non-rigid transform to temporally align and account for point cloud inconsistencies.…”
Section: Related Workmentioning
confidence: 99%
“…Alternative solutions rely on structure from motion (SfM) and multi-view stereo (MVS) to generate point clouds [16,17]. However, these methods are time-consuming and require several days to generate point clouds for a medium-sized urban environment [18].…”
Section: Introductionmentioning
confidence: 99%
“…Pang et al [5] created a 3D change map using a deep convolution network in addition to a graph-based and simultaneous segmentation method. Yew and Lee [45] applied a convolution network to balanced point clouds obtained from "structure from motion" of two times for urban CD. In another study, Yadav et al [46] used Lidar point clouds for 3D building CD in the Unet network.…”
Section: Introductionmentioning
confidence: 99%