Stacked convolutional auto-encoders for surface recognition based on 3d point cloud data

Maierdan Maimaitimin, Keigo Watanabe, Shoichi Maeyama

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

This paper addresses the problem of feature extraction for 3d point cloud data using a deep-structured auto-encoder. As one of the most focused research areas in human–robot interaction (HRI), the vision-based object recognition is very important. To recognize object using the most common geometry feature, surface condition that can be obtained from 3d point cloud data could decrease the error during the HRI. In this research, the surface normal vectors are used to convert 3D point cloud data to a surface-condition-feature map, and a sub-route stacked convolution auto-encoder (sCAE) is designed to classify the difference between the surfaces. The result of the trained filters and the classification of sCAE shows the surface-condition-feature and the specified sCAE are very effective in the variation of surface condition.

Original languageEnglish
Pages (from-to)259-264
Number of pages6
JournalArtificial Life and Robotics
Volume22
Issue number2
DOIs
Publication statusPublished - Jun 1 2017

Fingerprint

Convolution
Research
Object recognition
Feature extraction
Geometry

Keywords

  • 3D point cloud data
  • Convolution neural network
  • Sub-route convolution autoencoder
  • Surface-condition-feature

ASJC Scopus subject areas

  • Biochemistry, Genetics and Molecular Biology(all)
  • Artificial Intelligence

Cite this

Stacked convolutional auto-encoders for surface recognition based on 3d point cloud data. / Maimaitimin, Maierdan; Watanabe, Keigo; Maeyama, Shoichi.

In: Artificial Life and Robotics, Vol. 22, No. 2, 01.06.2017, p. 259-264.

Research output: Contribution to journalArticle

@article{bec9c382febc41a49144f86ffe424e2f,
title = "Stacked convolutional auto-encoders for surface recognition based on 3d point cloud data",
abstract = "This paper addresses the problem of feature extraction for 3d point cloud data using a deep-structured auto-encoder. As one of the most focused research areas in human–robot interaction (HRI), the vision-based object recognition is very important. To recognize object using the most common geometry feature, surface condition that can be obtained from 3d point cloud data could decrease the error during the HRI. In this research, the surface normal vectors are used to convert 3D point cloud data to a surface-condition-feature map, and a sub-route stacked convolution auto-encoder (sCAE) is designed to classify the difference between the surfaces. The result of the trained filters and the classification of sCAE shows the surface-condition-feature and the specified sCAE are very effective in the variation of surface condition.",
keywords = "3D point cloud data, Convolution neural network, Sub-route convolution autoencoder, Surface-condition-feature",
author = "Maierdan Maimaitimin and Keigo Watanabe and Shoichi Maeyama",
year = "2017",
month = "6",
day = "1",
doi = "10.1007/s10015-017-0350-9",
language = "English",
volume = "22",
pages = "259--264",
journal = "Artificial Life and Robotics",
issn = "1433-5298",
publisher = "Springer Japan",
number = "2",

}

TY - JOUR

T1 - Stacked convolutional auto-encoders for surface recognition based on 3d point cloud data

AU - Maimaitimin, Maierdan

AU - Watanabe, Keigo

AU - Maeyama, Shoichi

PY - 2017/6/1

Y1 - 2017/6/1

N2 - This paper addresses the problem of feature extraction for 3d point cloud data using a deep-structured auto-encoder. As one of the most focused research areas in human–robot interaction (HRI), the vision-based object recognition is very important. To recognize object using the most common geometry feature, surface condition that can be obtained from 3d point cloud data could decrease the error during the HRI. In this research, the surface normal vectors are used to convert 3D point cloud data to a surface-condition-feature map, and a sub-route stacked convolution auto-encoder (sCAE) is designed to classify the difference between the surfaces. The result of the trained filters and the classification of sCAE shows the surface-condition-feature and the specified sCAE are very effective in the variation of surface condition.

AB - This paper addresses the problem of feature extraction for 3d point cloud data using a deep-structured auto-encoder. As one of the most focused research areas in human–robot interaction (HRI), the vision-based object recognition is very important. To recognize object using the most common geometry feature, surface condition that can be obtained from 3d point cloud data could decrease the error during the HRI. In this research, the surface normal vectors are used to convert 3D point cloud data to a surface-condition-feature map, and a sub-route stacked convolution auto-encoder (sCAE) is designed to classify the difference between the surfaces. The result of the trained filters and the classification of sCAE shows the surface-condition-feature and the specified sCAE are very effective in the variation of surface condition.

KW - 3D point cloud data

KW - Convolution neural network

KW - Sub-route convolution autoencoder

KW - Surface-condition-feature

UR - http://www.scopus.com/inward/record.url?scp=85010966298&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85010966298&partnerID=8YFLogxK

U2 - 10.1007/s10015-017-0350-9

DO - 10.1007/s10015-017-0350-9

M3 - Article

VL - 22

SP - 259

EP - 264

JO - Artificial Life and Robotics

JF - Artificial Life and Robotics

SN - 1433-5298

IS - 2

ER -