DeepAI AI Chat
Log In Sign Up

Efficient moving point handling for incremental 3D manifold reconstruction

07/20/2015
by   Andrea Romanoni, et al.
Politecnico di Milano
0

As incremental Structure from Motion algorithms become effective, a good sparse point cloud representing the map of the scene becomes available frame-by-frame. From the 3D Delaunay triangulation of these points, state-of-the-art algorithms build a manifold rough model of the scene. These algorithms integrate incrementally new points to the 3D reconstruction only if their position estimate does not change. Indeed, whenever a point moves in a 3D Delaunay triangulation, for instance because its estimation gets refined, a set of tetrahedra have to be removed and replaced with new ones to maintain the Delaunay property; the management of the manifold reconstruction becomes thus complex and it entails a potentially big overhead. In this paper we investigate different approaches and we propose an efficient policy to deal with moving points in the manifold estimation process. We tested our approach with four sequences of the KITTI dataset and we show the effectiveness of our proposal in comparison with state-of-the-art approaches.

READ FULL TEXT
01/16/2018

Real-time CPU-based large-scale 3D mesh reconstruction

In Robotics, especially in this era of autonomous driving, mapping is on...
04/21/2016

Incremental Reconstruction of Urban Environments by Edge-Points Delaunay Triangulation

Urban reconstruction from a video captured by a surveying vehicle consti...
07/25/2022

Dynamic 3D Scene Analysis by Point Cloud Accumulation

Multi-beam LiDAR sensors, as used on autonomous vehicles and mobile robo...
07/20/2020

Approximating the Riemannian Metric from Point Clouds via Manifold Moving Least Squares

The approximation of both geodesic distances and shortest paths on point...
02/04/2019

Jumping Manifolds: Geometry Aware Dense Non-Rigid Structure from Motion

Given dense image feature correspondences of a non-rigidly moving object...
03/04/2019

Incremental Visual-Inertial 3D Mesh Generation with Structural Regularities

Visual-Inertial Odometry (VIO) algorithms typically rely on a point clou...
12/04/2020

Connecting 3-manifold triangulations with monotonic sequences of bistellar flips

A key result in computational 3-manifold topology is that any two triang...