3D object recognition using effective features selected by evaluating performance of discrimination

Shoichi Takei, Shuichi Akizuki, Manabu Hashimoto

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

1 Citation (Scopus)

Abstract

We propose a reliable 3D position and pose recognition method for complicated scenes including randomly stacked objects. Conventional methods use a small number of features selected by analyzing a target object model for recognition. The small number contributes to high-speed recognition, but actually the features include both 'true' and 'false' features. True features exist only in the target object and not in other parts, so they are valid for correct recognition purposes. On the other hand, false features exist in both the target object and in other parts, such as contacting areas caused by multiple objects. As a result of their matching incorrect parts, misrecognition may occur. To solve this problem, we propose a new method that uses effective features selected by analyzing not only the target object but also contacting areas caused by multiple objects. For predicting contacting areas, we generated very real input scenes by using 3D Computer Graphics (3D-CG) techniques and a physics-based simulator. Features that have high discrimination performance in the feature space are selected and used for the matching process. The method is robust to disturbances such as feature variability and achieves high feature separability; these enable it to achieve good discrimination performance. The method achieves reliable and fast object recognition by using a small number of effective features that have high discrimination performance. Experimental results show that the method's recognition success rate is from 33.6% to 92.9% higher than that of the Vector Pair Matching (VPM) method proposed by Akizuki et al. and that its processing time is within 1.46 seconds.

Original languageEnglish
Title of host publication2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages70-75
Number of pages6
ISBN (Electronic)9781479951994
DOIs
Publication statusPublished - 1997 Mar 19
Externally publishedYes
Event2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014 - Singapore, Singapore
Duration: 2014 Dec 102014 Dec 12

Other

Other2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014
CountrySingapore
CitySingapore
Period14/12/1014/12/12

Fingerprint

Object recognition
Computer graphics
Physics
Simulators
Processing

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Human-Computer Interaction
  • Artificial Intelligence
  • Control and Systems Engineering

Cite this

Takei, S., Akizuki, S., & Hashimoto, M. (1997). 3D object recognition using effective features selected by evaluating performance of discrimination. In 2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014 (pp. 70-75). [7064282] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICARCV.2014.7064282

3D object recognition using effective features selected by evaluating performance of discrimination. / Takei, Shoichi; Akizuki, Shuichi; Hashimoto, Manabu.

2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014. Institute of Electrical and Electronics Engineers Inc., 1997. p. 70-75 7064282.

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

Takei, S, Akizuki, S & Hashimoto, M 1997, 3D object recognition using effective features selected by evaluating performance of discrimination. in 2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014., 7064282, Institute of Electrical and Electronics Engineers Inc., pp. 70-75, 2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014, Singapore, Singapore, 14/12/10. https://doi.org/10.1109/ICARCV.2014.7064282
Takei S, Akizuki S, Hashimoto M. 3D object recognition using effective features selected by evaluating performance of discrimination. In 2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014. Institute of Electrical and Electronics Engineers Inc. 1997. p. 70-75. 7064282 https://doi.org/10.1109/ICARCV.2014.7064282
Takei, Shoichi ; Akizuki, Shuichi ; Hashimoto, Manabu. / 3D object recognition using effective features selected by evaluating performance of discrimination. 2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014. Institute of Electrical and Electronics Engineers Inc., 1997. pp. 70-75
@inproceedings{66cd551ff8b64387bddef54383fca78c,
title = "3D object recognition using effective features selected by evaluating performance of discrimination",
abstract = "We propose a reliable 3D position and pose recognition method for complicated scenes including randomly stacked objects. Conventional methods use a small number of features selected by analyzing a target object model for recognition. The small number contributes to high-speed recognition, but actually the features include both 'true' and 'false' features. True features exist only in the target object and not in other parts, so they are valid for correct recognition purposes. On the other hand, false features exist in both the target object and in other parts, such as contacting areas caused by multiple objects. As a result of their matching incorrect parts, misrecognition may occur. To solve this problem, we propose a new method that uses effective features selected by analyzing not only the target object but also contacting areas caused by multiple objects. For predicting contacting areas, we generated very real input scenes by using 3D Computer Graphics (3D-CG) techniques and a physics-based simulator. Features that have high discrimination performance in the feature space are selected and used for the matching process. The method is robust to disturbances such as feature variability and achieves high feature separability; these enable it to achieve good discrimination performance. The method achieves reliable and fast object recognition by using a small number of effective features that have high discrimination performance. Experimental results show that the method's recognition success rate is from 33.6{\%} to 92.9{\%} higher than that of the Vector Pair Matching (VPM) method proposed by Akizuki et al. and that its processing time is within 1.46 seconds.",
author = "Shoichi Takei and Shuichi Akizuki and Manabu Hashimoto",
year = "1997",
month = "3",
day = "19",
doi = "10.1109/ICARCV.2014.7064282",
language = "English",
pages = "70--75",
booktitle = "2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - 3D object recognition using effective features selected by evaluating performance of discrimination

AU - Takei, Shoichi

AU - Akizuki, Shuichi

AU - Hashimoto, Manabu

PY - 1997/3/19

Y1 - 1997/3/19

N2 - We propose a reliable 3D position and pose recognition method for complicated scenes including randomly stacked objects. Conventional methods use a small number of features selected by analyzing a target object model for recognition. The small number contributes to high-speed recognition, but actually the features include both 'true' and 'false' features. True features exist only in the target object and not in other parts, so they are valid for correct recognition purposes. On the other hand, false features exist in both the target object and in other parts, such as contacting areas caused by multiple objects. As a result of their matching incorrect parts, misrecognition may occur. To solve this problem, we propose a new method that uses effective features selected by analyzing not only the target object but also contacting areas caused by multiple objects. For predicting contacting areas, we generated very real input scenes by using 3D Computer Graphics (3D-CG) techniques and a physics-based simulator. Features that have high discrimination performance in the feature space are selected and used for the matching process. The method is robust to disturbances such as feature variability and achieves high feature separability; these enable it to achieve good discrimination performance. The method achieves reliable and fast object recognition by using a small number of effective features that have high discrimination performance. Experimental results show that the method's recognition success rate is from 33.6% to 92.9% higher than that of the Vector Pair Matching (VPM) method proposed by Akizuki et al. and that its processing time is within 1.46 seconds.

AB - We propose a reliable 3D position and pose recognition method for complicated scenes including randomly stacked objects. Conventional methods use a small number of features selected by analyzing a target object model for recognition. The small number contributes to high-speed recognition, but actually the features include both 'true' and 'false' features. True features exist only in the target object and not in other parts, so they are valid for correct recognition purposes. On the other hand, false features exist in both the target object and in other parts, such as contacting areas caused by multiple objects. As a result of their matching incorrect parts, misrecognition may occur. To solve this problem, we propose a new method that uses effective features selected by analyzing not only the target object but also contacting areas caused by multiple objects. For predicting contacting areas, we generated very real input scenes by using 3D Computer Graphics (3D-CG) techniques and a physics-based simulator. Features that have high discrimination performance in the feature space are selected and used for the matching process. The method is robust to disturbances such as feature variability and achieves high feature separability; these enable it to achieve good discrimination performance. The method achieves reliable and fast object recognition by using a small number of effective features that have high discrimination performance. Experimental results show that the method's recognition success rate is from 33.6% to 92.9% higher than that of the Vector Pair Matching (VPM) method proposed by Akizuki et al. and that its processing time is within 1.46 seconds.

UR - http://www.scopus.com/inward/record.url?scp=84949924646&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84949924646&partnerID=8YFLogxK

U2 - 10.1109/ICARCV.2014.7064282

DO - 10.1109/ICARCV.2014.7064282

M3 - Conference contribution

SP - 70

EP - 75

BT - 2014 13th International Conference on Control Automation Robotics and Vision, ICARCV 2014

PB - Institute of Electrical and Electronics Engineers Inc.

ER -