Reliable object recognition using 3-D feature points for minimizing its mismatching

Masanobu Nagase, Shuichi Akizuki, Manabu Hashimoto

Research output: Contribution to journalArticle

Abstract

This paper introduces a reliable 3-D object recognition method which can statistically minimize its mismatching. This method basically uses a 3-D object model which is represented as a set of feature points with 3-D coordinate. Each feature point also has an attribute value about local shape around the point. The attribute value is represented as an orientation histogram of normal vector, which can be calculated by using several neighboring feature points around each point. Here, this attribute value means its local shape. By estimating relative similarity of two points of all possible combination in the model, the distinctiveness of each point is defined. In the proposed method, only a small number of distinctive feature points are selected and used for matching with all feature points extracted from an acquired range image. Finally, position and pose of the target object can be estimated from some correctly matched points. Experimental results using real scene have proved that recognition rate of this method is 93.8%, which is 42.2% higher than the conventional Spin Image method. Also, computing time is about nine times faster than that.

Original languageEnglish
Pages (from-to)1058-1062
Number of pages5
JournalSeimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering
Volume79
Issue number11
Publication statusPublished - 2013 Nov
Externally publishedYes

Fingerprint

Object recognition

Keywords

  • 3-D descriptor
  • 3-D feature point matching
  • Bin-picking
  • Object recognition
  • Point cloud data
  • Robot vision

ASJC Scopus subject areas

  • Mechanical Engineering

Cite this

Reliable object recognition using 3-D feature points for minimizing its mismatching. / Nagase, Masanobu; Akizuki, Shuichi; Hashimoto, Manabu.

In: Seimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering, Vol. 79, No. 11, 11.2013, p. 1058-1062.

Research output: Contribution to journalArticle

@article{14956c473dd247f9b3e289168ee9b1d1,
title = "Reliable object recognition using 3-D feature points for minimizing its mismatching",
abstract = "This paper introduces a reliable 3-D object recognition method which can statistically minimize its mismatching. This method basically uses a 3-D object model which is represented as a set of feature points with 3-D coordinate. Each feature point also has an attribute value about local shape around the point. The attribute value is represented as an orientation histogram of normal vector, which can be calculated by using several neighboring feature points around each point. Here, this attribute value means its local shape. By estimating relative similarity of two points of all possible combination in the model, the distinctiveness of each point is defined. In the proposed method, only a small number of distinctive feature points are selected and used for matching with all feature points extracted from an acquired range image. Finally, position and pose of the target object can be estimated from some correctly matched points. Experimental results using real scene have proved that recognition rate of this method is 93.8{\%}, which is 42.2{\%} higher than the conventional Spin Image method. Also, computing time is about nine times faster than that.",
keywords = "3-D descriptor, 3-D feature point matching, Bin-picking, Object recognition, Point cloud data, Robot vision",
author = "Masanobu Nagase and Shuichi Akizuki and Manabu Hashimoto",
year = "2013",
month = "11",
language = "English",
volume = "79",
pages = "1058--1062",
journal = "Seimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering",
issn = "0912-0289",
publisher = "Japan Society for Precision Engineering",
number = "11",

}

TY - JOUR

T1 - Reliable object recognition using 3-D feature points for minimizing its mismatching

AU - Nagase, Masanobu

AU - Akizuki, Shuichi

AU - Hashimoto, Manabu

PY - 2013/11

Y1 - 2013/11

N2 - This paper introduces a reliable 3-D object recognition method which can statistically minimize its mismatching. This method basically uses a 3-D object model which is represented as a set of feature points with 3-D coordinate. Each feature point also has an attribute value about local shape around the point. The attribute value is represented as an orientation histogram of normal vector, which can be calculated by using several neighboring feature points around each point. Here, this attribute value means its local shape. By estimating relative similarity of two points of all possible combination in the model, the distinctiveness of each point is defined. In the proposed method, only a small number of distinctive feature points are selected and used for matching with all feature points extracted from an acquired range image. Finally, position and pose of the target object can be estimated from some correctly matched points. Experimental results using real scene have proved that recognition rate of this method is 93.8%, which is 42.2% higher than the conventional Spin Image method. Also, computing time is about nine times faster than that.

AB - This paper introduces a reliable 3-D object recognition method which can statistically minimize its mismatching. This method basically uses a 3-D object model which is represented as a set of feature points with 3-D coordinate. Each feature point also has an attribute value about local shape around the point. The attribute value is represented as an orientation histogram of normal vector, which can be calculated by using several neighboring feature points around each point. Here, this attribute value means its local shape. By estimating relative similarity of two points of all possible combination in the model, the distinctiveness of each point is defined. In the proposed method, only a small number of distinctive feature points are selected and used for matching with all feature points extracted from an acquired range image. Finally, position and pose of the target object can be estimated from some correctly matched points. Experimental results using real scene have proved that recognition rate of this method is 93.8%, which is 42.2% higher than the conventional Spin Image method. Also, computing time is about nine times faster than that.

KW - 3-D descriptor

KW - 3-D feature point matching

KW - Bin-picking

KW - Object recognition

KW - Point cloud data

KW - Robot vision

UR - http://www.scopus.com/inward/record.url?scp=84888251624&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84888251624&partnerID=8YFLogxK

M3 - Article

VL - 79

SP - 1058

EP - 1062

JO - Seimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering

JF - Seimitsu Kogaku Kaishi/Journal of the Japan Society for Precision Engineering

SN - 0912-0289

IS - 11

ER -