SLAM Using 3D reconstruction via a visual RGB and RGB-D sensory input

This paper investigates simultaneous localization and mapping (SLAM) problem by exploiting the Microsoft Kinect™ sensor array and an autonomous mobile robot capable of self-localization. The combination of them covers the major features of SLAM including mapping, sensing, locating, and modeling. The...

Full description

Bibliographic Details
Main Authors: Wurdemann, H., Georgiou, E., Cui, Lei, Dai, J.
Other Authors: Primo Zingaretti
Format: Conference Paper
Published: ASME Press 2011
Subjects:
Online Access:http://hdl.handle.net/20.500.11937/22225
_version_ 1848750811065090048
author Wurdemann, H.
Georgiou, E.
Cui, Lei
Dai, J.
author2 Primo Zingaretti
author_facet Primo Zingaretti
Wurdemann, H.
Georgiou, E.
Cui, Lei
Dai, J.
author_sort Wurdemann, H.
building Curtin Institutional Repository
collection Online Access
description This paper investigates simultaneous localization and mapping (SLAM) problem by exploiting the Microsoft Kinect™ sensor array and an autonomous mobile robot capable of self-localization. The combination of them covers the major features of SLAM including mapping, sensing, locating, and modeling. The Kinect™ sensor array provides a dual camera output of RGB, using a CMOS camera, and RGB-D, using a depth camera. The sensors will be mounted on the KCLBOT, an autonomous nonholonomic two wheel maneuverable mobile robot. The mobile robot platform has the ability to self-localize and preform navigation maneuvers to traverse to set target points using intelligent processes. The target point for this operation is a fixed coordinate position, which will be the goal for the mobile robot to reach, taking into consideration the obstacles in the environment which will be represented in a 3D spatial model. Extracting the images from the sensor after a calibration routine, a 3D reconstruction of the traversable environment is produced for the mobile robot to navigate. Using the constructed 3D model the autonomous mobile robot follows a polynomial-based nonholonomic trajectory with obstacle avoidance. The experimental results demonstrate the cost effectiveness of this off the shelf sensor array. The results show the effectiveness to produce a 3D reconstruction of an environment and the feasibility of using the Microsoft Kinect™ sensor for mapping, sensing, locating, and modeling, that enables the implementation of SLAM on this type of platform.
first_indexed 2025-11-14T07:42:45Z
format Conference Paper
id curtin-20.500.11937-22225
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T07:42:45Z
publishDate 2011
publisher ASME Press
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-222252017-09-13T13:50:22Z SLAM Using 3D reconstruction via a visual RGB and RGB-D sensory input Wurdemann, H. Georgiou, E. Cui, Lei Dai, J. Primo Zingaretti Kinect 3D reconstruction SLAM This paper investigates simultaneous localization and mapping (SLAM) problem by exploiting the Microsoft Kinect™ sensor array and an autonomous mobile robot capable of self-localization. The combination of them covers the major features of SLAM including mapping, sensing, locating, and modeling. The Kinect™ sensor array provides a dual camera output of RGB, using a CMOS camera, and RGB-D, using a depth camera. The sensors will be mounted on the KCLBOT, an autonomous nonholonomic two wheel maneuverable mobile robot. The mobile robot platform has the ability to self-localize and preform navigation maneuvers to traverse to set target points using intelligent processes. The target point for this operation is a fixed coordinate position, which will be the goal for the mobile robot to reach, taking into consideration the obstacles in the environment which will be represented in a 3D spatial model. Extracting the images from the sensor after a calibration routine, a 3D reconstruction of the traversable environment is produced for the mobile robot to navigate. Using the constructed 3D model the autonomous mobile robot follows a polynomial-based nonholonomic trajectory with obstacle avoidance. The experimental results demonstrate the cost effectiveness of this off the shelf sensor array. The results show the effectiveness to produce a 3D reconstruction of an environment and the feasibility of using the Microsoft Kinect™ sensor for mapping, sensing, locating, and modeling, that enables the implementation of SLAM on this type of platform. 2011 Conference Paper http://hdl.handle.net/20.500.11937/22225 10.1115/DETC2011-47735 ASME Press restricted
spellingShingle Kinect
3D reconstruction
SLAM
Wurdemann, H.
Georgiou, E.
Cui, Lei
Dai, J.
SLAM Using 3D reconstruction via a visual RGB and RGB-D sensory input
title SLAM Using 3D reconstruction via a visual RGB and RGB-D sensory input
title_full SLAM Using 3D reconstruction via a visual RGB and RGB-D sensory input
title_fullStr SLAM Using 3D reconstruction via a visual RGB and RGB-D sensory input
title_full_unstemmed SLAM Using 3D reconstruction via a visual RGB and RGB-D sensory input
title_short SLAM Using 3D reconstruction via a visual RGB and RGB-D sensory input
title_sort slam using 3d reconstruction via a visual rgb and rgb-d sensory input
topic Kinect
3D reconstruction
SLAM
url http://hdl.handle.net/20.500.11937/22225