Surface-common-feature descriptor of point cloud data for deep learning

Maierdan Maimaitimin, Keigo Watanabe, Shoichi Maeyama

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper addresses the problem of feature extraction for 3d point cloud data by using autoencoder. Deep learning is one of the most active fields of artificial intelligence, especially in a variety of visual applications, such as image classification and object recognition. However it has not been successfully applied on 3d point cloud data. In this paper, a new method of analyzing the point cloud data is proposed. The method aims to convert the point cloud data to a surface-condition-feature map, which is very effective and useful in pre-training by autoencoder. The surface-condition-features in this paper are defined as upward inclined, downward inclined, upward curved, downward curved, edge and flat, where those features are converted from surface normal vectors.

Original languageEnglish
Title of host publication2016 IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages525-529
Number of pages5
ISBN (Electronic)9781509023943
DOIs
Publication statusPublished - Sep 1 2016
Event13th IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016 - Harbin, Heilongjiang, China
Duration: Aug 7 2016Aug 10 2016

Other

Other13th IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016
CountryChina
CityHarbin, Heilongjiang
Period8/7/168/10/16

Fingerprint

Image classification
Object recognition
Artificial intelligence
Feature extraction
Deep learning

ASJC Scopus subject areas

  • Mechanical Engineering
  • Artificial Intelligence
  • Computer Science Applications
  • Software

Cite this

Maimaitimin, M., Watanabe, K., & Maeyama, S. (2016). Surface-common-feature descriptor of point cloud data for deep learning. In 2016 IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016 (pp. 525-529). [7558618] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICMA.2016.7558618

Surface-common-feature descriptor of point cloud data for deep learning. / Maimaitimin, Maierdan; Watanabe, Keigo; Maeyama, Shoichi.

2016 IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016. Institute of Electrical and Electronics Engineers Inc., 2016. p. 525-529 7558618.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Maimaitimin, M, Watanabe, K & Maeyama, S 2016, Surface-common-feature descriptor of point cloud data for deep learning. in 2016 IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016., 7558618, Institute of Electrical and Electronics Engineers Inc., pp. 525-529, 13th IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016, Harbin, Heilongjiang, China, 8/7/16. https://doi.org/10.1109/ICMA.2016.7558618
Maimaitimin M, Watanabe K, Maeyama S. Surface-common-feature descriptor of point cloud data for deep learning. In 2016 IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016. Institute of Electrical and Electronics Engineers Inc. 2016. p. 525-529. 7558618 https://doi.org/10.1109/ICMA.2016.7558618
Maimaitimin, Maierdan ; Watanabe, Keigo ; Maeyama, Shoichi. / Surface-common-feature descriptor of point cloud data for deep learning. 2016 IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016. Institute of Electrical and Electronics Engineers Inc., 2016. pp. 525-529
@inproceedings{37bc649515274519bd7a38a88a38188b,
title = "Surface-common-feature descriptor of point cloud data for deep learning",
abstract = "This paper addresses the problem of feature extraction for 3d point cloud data by using autoencoder. Deep learning is one of the most active fields of artificial intelligence, especially in a variety of visual applications, such as image classification and object recognition. However it has not been successfully applied on 3d point cloud data. In this paper, a new method of analyzing the point cloud data is proposed. The method aims to convert the point cloud data to a surface-condition-feature map, which is very effective and useful in pre-training by autoencoder. The surface-condition-features in this paper are defined as upward inclined, downward inclined, upward curved, downward curved, edge and flat, where those features are converted from surface normal vectors.",
author = "Maierdan Maimaitimin and Keigo Watanabe and Shoichi Maeyama",
year = "2016",
month = "9",
day = "1",
doi = "10.1109/ICMA.2016.7558618",
language = "English",
pages = "525--529",
booktitle = "2016 IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Surface-common-feature descriptor of point cloud data for deep learning

AU - Maimaitimin, Maierdan

AU - Watanabe, Keigo

AU - Maeyama, Shoichi

PY - 2016/9/1

Y1 - 2016/9/1

N2 - This paper addresses the problem of feature extraction for 3d point cloud data by using autoencoder. Deep learning is one of the most active fields of artificial intelligence, especially in a variety of visual applications, such as image classification and object recognition. However it has not been successfully applied on 3d point cloud data. In this paper, a new method of analyzing the point cloud data is proposed. The method aims to convert the point cloud data to a surface-condition-feature map, which is very effective and useful in pre-training by autoencoder. The surface-condition-features in this paper are defined as upward inclined, downward inclined, upward curved, downward curved, edge and flat, where those features are converted from surface normal vectors.

AB - This paper addresses the problem of feature extraction for 3d point cloud data by using autoencoder. Deep learning is one of the most active fields of artificial intelligence, especially in a variety of visual applications, such as image classification and object recognition. However it has not been successfully applied on 3d point cloud data. In this paper, a new method of analyzing the point cloud data is proposed. The method aims to convert the point cloud data to a surface-condition-feature map, which is very effective and useful in pre-training by autoencoder. The surface-condition-features in this paper are defined as upward inclined, downward inclined, upward curved, downward curved, edge and flat, where those features are converted from surface normal vectors.

UR - http://www.scopus.com/inward/record.url?scp=84991222063&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84991222063&partnerID=8YFLogxK

U2 - 10.1109/ICMA.2016.7558618

DO - 10.1109/ICMA.2016.7558618

M3 - Conference contribution

SP - 525

EP - 529

BT - 2016 IEEE International Conference on Mechatronics and Automation, IEEE ICMA 2016

PB - Institute of Electrical and Electronics Engineers Inc.

ER -