Virtual object for evaluating adaptable K-nearest neighbor method solving various conditions of object recognition

Wittayathawon Kanlaya, Le Dung, Makoto Mizukawa

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

2 Citations (Scopus)

Abstract

In order for robots to be able to manipulate the proper objects, robots firstly need visual ability to precisely recognize and identify objects. One of the most basic problems with robot vision is that environments can change under various weather conditions (various illuminations). Furthermore, each object's category consists of many objects with various poses. In order to obtain the best performance in term of accuracy and efficiency, we compared three feature extraction approaches that have been widely used to solve this problem: Principal Components Analysis (PCA), Linear Discriminant Analysis (LDA), and contour matching with Log Polar Histogram (LPH). We also introduced an improved algorithm called Adaptable K-Nearest Neighbor (AK-NN) that allows the object recognition system to use an automatic adaptable K value to improve the accuracy of classification. To evaluate the object recognition system, we generated virtual objects with various conditions for realistic testing.

Original languageEnglish
Title of host publicationICCAS-SICE 2009 - ICROS-SICE International Joint Conference 2009, Proceedings
Pages4338-4342
Number of pages5
Publication statusPublished - 2009
EventICROS-SICE International Joint Conference 2009, ICCAS-SICE 2009 - Fukuoka
Duration: 2009 Aug 182009 Aug 21

Other

OtherICROS-SICE International Joint Conference 2009, ICCAS-SICE 2009
CityFukuoka
Period09/8/1809/8/21

Fingerprint

Object recognition
Robots
Discriminant analysis
Principal component analysis
Computer vision
Feature extraction
Lighting
Testing

Keywords

  • Object recognition
  • Robot vision
  • Virtual object

ASJC Scopus subject areas

  • Information Systems
  • Control and Systems Engineering
  • Industrial and Manufacturing Engineering

Cite this

Kanlaya, W., Dung, L., & Mizukawa, M. (2009). Virtual object for evaluating adaptable K-nearest neighbor method solving various conditions of object recognition. In ICCAS-SICE 2009 - ICROS-SICE International Joint Conference 2009, Proceedings (pp. 4338-4342). [5332838]

Virtual object for evaluating adaptable K-nearest neighbor method solving various conditions of object recognition. / Kanlaya, Wittayathawon; Dung, Le; Mizukawa, Makoto.

ICCAS-SICE 2009 - ICROS-SICE International Joint Conference 2009, Proceedings. 2009. p. 4338-4342 5332838.

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

Kanlaya, W, Dung, L & Mizukawa, M 2009, Virtual object for evaluating adaptable K-nearest neighbor method solving various conditions of object recognition. in ICCAS-SICE 2009 - ICROS-SICE International Joint Conference 2009, Proceedings., 5332838, pp. 4338-4342, ICROS-SICE International Joint Conference 2009, ICCAS-SICE 2009, Fukuoka, 09/8/18.
Kanlaya W, Dung L, Mizukawa M. Virtual object for evaluating adaptable K-nearest neighbor method solving various conditions of object recognition. In ICCAS-SICE 2009 - ICROS-SICE International Joint Conference 2009, Proceedings. 2009. p. 4338-4342. 5332838
Kanlaya, Wittayathawon ; Dung, Le ; Mizukawa, Makoto. / Virtual object for evaluating adaptable K-nearest neighbor method solving various conditions of object recognition. ICCAS-SICE 2009 - ICROS-SICE International Joint Conference 2009, Proceedings. 2009. pp. 4338-4342
@inproceedings{d227bcee0cdc4264b465e627b3a544db,
title = "Virtual object for evaluating adaptable K-nearest neighbor method solving various conditions of object recognition",
abstract = "In order for robots to be able to manipulate the proper objects, robots firstly need visual ability to precisely recognize and identify objects. One of the most basic problems with robot vision is that environments can change under various weather conditions (various illuminations). Furthermore, each object's category consists of many objects with various poses. In order to obtain the best performance in term of accuracy and efficiency, we compared three feature extraction approaches that have been widely used to solve this problem: Principal Components Analysis (PCA), Linear Discriminant Analysis (LDA), and contour matching with Log Polar Histogram (LPH). We also introduced an improved algorithm called Adaptable K-Nearest Neighbor (AK-NN) that allows the object recognition system to use an automatic adaptable K value to improve the accuracy of classification. To evaluate the object recognition system, we generated virtual objects with various conditions for realistic testing.",
keywords = "Object recognition, Robot vision, Virtual object",
author = "Wittayathawon Kanlaya and Le Dung and Makoto Mizukawa",
year = "2009",
language = "English",
isbn = "9784907764333",
pages = "4338--4342",
booktitle = "ICCAS-SICE 2009 - ICROS-SICE International Joint Conference 2009, Proceedings",

}

TY - GEN

T1 - Virtual object for evaluating adaptable K-nearest neighbor method solving various conditions of object recognition

AU - Kanlaya, Wittayathawon

AU - Dung, Le

AU - Mizukawa, Makoto

PY - 2009

Y1 - 2009

N2 - In order for robots to be able to manipulate the proper objects, robots firstly need visual ability to precisely recognize and identify objects. One of the most basic problems with robot vision is that environments can change under various weather conditions (various illuminations). Furthermore, each object's category consists of many objects with various poses. In order to obtain the best performance in term of accuracy and efficiency, we compared three feature extraction approaches that have been widely used to solve this problem: Principal Components Analysis (PCA), Linear Discriminant Analysis (LDA), and contour matching with Log Polar Histogram (LPH). We also introduced an improved algorithm called Adaptable K-Nearest Neighbor (AK-NN) that allows the object recognition system to use an automatic adaptable K value to improve the accuracy of classification. To evaluate the object recognition system, we generated virtual objects with various conditions for realistic testing.

AB - In order for robots to be able to manipulate the proper objects, robots firstly need visual ability to precisely recognize and identify objects. One of the most basic problems with robot vision is that environments can change under various weather conditions (various illuminations). Furthermore, each object's category consists of many objects with various poses. In order to obtain the best performance in term of accuracy and efficiency, we compared three feature extraction approaches that have been widely used to solve this problem: Principal Components Analysis (PCA), Linear Discriminant Analysis (LDA), and contour matching with Log Polar Histogram (LPH). We also introduced an improved algorithm called Adaptable K-Nearest Neighbor (AK-NN) that allows the object recognition system to use an automatic adaptable K value to improve the accuracy of classification. To evaluate the object recognition system, we generated virtual objects with various conditions for realistic testing.

KW - Object recognition

KW - Robot vision

KW - Virtual object

UR - http://www.scopus.com/inward/record.url?scp=77951125908&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77951125908&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9784907764333

SP - 4338

EP - 4342

BT - ICCAS-SICE 2009 - ICROS-SICE International Joint Conference 2009, Proceedings

ER -