Terrain Mapping and Obstacle Detection Using Gaussian Processes

Morten Kjærgaard; Alessandro Salvatore Massaro; Enis Bayramoglu; Kjeld Jensen

doi:10.1109/ICMLA.2011.137

Terrain Mapping and Obstacle Detection Using Gaussian Processes

Morten Kjærgaard, Alessandro Salvatore Massaro, Enis Bayramoglu, Kjeld Jensen

University of Southern Denmark

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Abstract

In this paper we consider a probabilistic method for extracting terrain maps from a scene and use the information to detect potential navigation obstacles within it. The method uses Gaussian process regression (GPR) to predict an estimate function and its relative uncertainty. To test the new methods, we have arranged two setups: an artificial flat surface with an object in front of the sensors and an outdoor unstructured terrain. Two sensor types have been used to determine the point cloud fed to the system: a 3D laser scanner and a stereo camera pair. The results from both sensor systems show that the estimated maps follow the terrain shape, while protrusions are identified and may be isolated as potential obstacles. Representing the data with a covariance function allows a dramatic reduction of the amount of data to process, while maintaining the statistical properties of the measured and interpolated features.

Original language	English
Title of host publication	Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011
Number of pages	6
Publication date	2011
Pages	118-123
ISBN (Print)	9780769546070
DOIs	https://doi.org/10.1109/ICMLA.2011.137
Publication status	Published - 2011
Event	10th International Conference on Machine Learning and Applications (ICMLA 2011) - Honolulu, Hawaii, United States Duration: 18 Dec 2011 → 21 Dec 2011 Conference number: 10

Conference

Conference	10th International Conference on Machine Learning and Applications (ICMLA 2011)
Number	10
Country/Territory	United States
City	Honolulu, Hawaii
Period	18/12/2011 → 21/12/2011

Access to Document

10.1109/ICMLA.2011.137

http://www.icmla-conference.org/icmla11/

OpenUrl availability

Full text

Cite this

@inproceedings{dbe888ed359a4a7ab5f667daa7474aa3,

title = "Terrain Mapping and Obstacle Detection Using Gaussian Processes",

abstract = "In this paper we consider a probabilistic method for extracting terrain maps from a scene and use the information to detect potential navigation obstacles within it. The method uses Gaussian process regression (GPR) to predict an estimate function and its relative uncertainty. To test the new methods, we have arranged two setups: an artificial flat surface with an object in front of the sensors and an outdoor unstructured terrain. Two sensor types have been used to determine the point cloud fed to the system: a 3D laser scanner and a stereo camera pair. The results from both sensor systems show that the estimated maps follow the terrain shape, while protrusions are identified and may be isolated as potential obstacles. Representing the data with a covariance function allows a dramatic reduction of the amount of data to process, while maintaining the statistical properties of the measured and interpolated features.",

author = "Morten Kj{\ae}rgaard and Massaro, {Alessandro Salvatore} and Enis Bayramoglu and Kjeld Jensen",

year = "2011",

doi = "10.1109/ICMLA.2011.137",

language = "English",

isbn = "9780769546070",

pages = "118--123",

booktitle = "Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011",

note = "10th International Conference on Machine Learning and Applications (ICMLA 2011), ICMLA ; Conference date: 18-12-2011 Through 21-12-2011",

}

Kjærgaard, M, Massaro, AS, Bayramoglu, E & Jensen, K 2011, Terrain Mapping and Obstacle Detection Using Gaussian Processes. in Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011. pp. 118-123, 10th International Conference on Machine Learning and Applications (ICMLA 2011), Honolulu, Hawaii, United States, 18/12/2011. https://doi.org/10.1109/ICMLA.2011.137

Terrain Mapping and Obstacle Detection Using Gaussian Processes. / Kjærgaard, Morten; Massaro, Alessandro Salvatore; Bayramoglu, Enis et al.
Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011. 2011. p. 118-123.

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

TY - GEN

T1 - Terrain Mapping and Obstacle Detection Using Gaussian Processes

AU - Kjærgaard, Morten

AU - Massaro, Alessandro Salvatore

AU - Bayramoglu, Enis

AU - Jensen, Kjeld

N1 - Conference code: 10

PY - 2011

Y1 - 2011

N2 - In this paper we consider a probabilistic method for extracting terrain maps from a scene and use the information to detect potential navigation obstacles within it. The method uses Gaussian process regression (GPR) to predict an estimate function and its relative uncertainty. To test the new methods, we have arranged two setups: an artificial flat surface with an object in front of the sensors and an outdoor unstructured terrain. Two sensor types have been used to determine the point cloud fed to the system: a 3D laser scanner and a stereo camera pair. The results from both sensor systems show that the estimated maps follow the terrain shape, while protrusions are identified and may be isolated as potential obstacles. Representing the data with a covariance function allows a dramatic reduction of the amount of data to process, while maintaining the statistical properties of the measured and interpolated features.

AB - In this paper we consider a probabilistic method for extracting terrain maps from a scene and use the information to detect potential navigation obstacles within it. The method uses Gaussian process regression (GPR) to predict an estimate function and its relative uncertainty. To test the new methods, we have arranged two setups: an artificial flat surface with an object in front of the sensors and an outdoor unstructured terrain. Two sensor types have been used to determine the point cloud fed to the system: a 3D laser scanner and a stereo camera pair. The results from both sensor systems show that the estimated maps follow the terrain shape, while protrusions are identified and may be isolated as potential obstacles. Representing the data with a covariance function allows a dramatic reduction of the amount of data to process, while maintaining the statistical properties of the measured and interpolated features.

U2 - 10.1109/ICMLA.2011.137

DO - 10.1109/ICMLA.2011.137

M3 - Article in proceedings

SN - 9780769546070

SP - 118

EP - 123

BT - Proceedings - 10th International Conference on Machine Learning and Applications, ICMLA 2011

T2 - 10th International Conference on Machine Learning and Applications (ICMLA 2011)

Y2 - 18 December 2011 through 21 December 2011

ER -

Terrain Mapping and Obstacle Detection Using Gaussian Processes

Abstract

Conference

Access to Document

OpenUrl availability

Fingerprint

Cite this