Hilbert scan based bag-of-features for image retrieval

Pengyi Hao, Seiichiro Kamata

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Generally, two problems of bag-of-features in image retrieval are still considered unsolved: one is that spatial information about descriptors is not employed well, which affects the accuracy of retrieval; the other is that the trade-off between vocabulary size and good precision, which decides the storage and retrieval performance. In this paper, we propose a novel approach called Hilbert scan based bag-of-features (HS-BoF) for image retrieval. Firstly, Hilbert scan based tree representation (HSBT) is studied, which is built based on the local descriptors while spatial relationships are added into the nodes by a novel grouping rule, resulting of a tree structure for each image. Further, we give two ways of codebook production based on HSBT: multi-layer codebook and multi-size codebook. Owing to the properties of Hilbert scanning and the merits of our grouping method, sub-regions of the tree are not only flexible to the distribution of local patches but also have hierarchical relations. Extensive experiments on caltech-256, 13-scene and 1 million ImageNet images show that HS-BoF obtains higher accuracy with less memory usage.

Original languageEnglish
Pages (from-to)1260-1268
Number of pages9
JournalIEICE Transactions on Information and Systems
VolumeE94-D
Issue number6
DOIs
Publication statusPublished - 2011 Jun

Fingerprint

Image retrieval
Scanning
Data storage equipment
Experiments

Keywords

  • Bag-of-features
  • Feature representation
  • Hilbert scanning
  • Image search

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Software
  • Artificial Intelligence
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition

Cite this

Hilbert scan based bag-of-features for image retrieval. / Hao, Pengyi; Kamata, Seiichiro.

In: IEICE Transactions on Information and Systems, Vol. E94-D, No. 6, 06.2011, p. 1260-1268.

Research output: Contribution to journalArticle

@article{862dc2b625054127af3be1566d132b68,
title = "Hilbert scan based bag-of-features for image retrieval",
abstract = "Generally, two problems of bag-of-features in image retrieval are still considered unsolved: one is that spatial information about descriptors is not employed well, which affects the accuracy of retrieval; the other is that the trade-off between vocabulary size and good precision, which decides the storage and retrieval performance. In this paper, we propose a novel approach called Hilbert scan based bag-of-features (HS-BoF) for image retrieval. Firstly, Hilbert scan based tree representation (HSBT) is studied, which is built based on the local descriptors while spatial relationships are added into the nodes by a novel grouping rule, resulting of a tree structure for each image. Further, we give two ways of codebook production based on HSBT: multi-layer codebook and multi-size codebook. Owing to the properties of Hilbert scanning and the merits of our grouping method, sub-regions of the tree are not only flexible to the distribution of local patches but also have hierarchical relations. Extensive experiments on caltech-256, 13-scene and 1 million ImageNet images show that HS-BoF obtains higher accuracy with less memory usage.",
keywords = "Bag-of-features, Feature representation, Hilbert scanning, Image search",
author = "Pengyi Hao and Seiichiro Kamata",
year = "2011",
month = "6",
doi = "10.1587/transinf.E94.D.1260",
language = "English",
volume = "E94-D",
pages = "1260--1268",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "6",

}

TY - JOUR

T1 - Hilbert scan based bag-of-features for image retrieval

AU - Hao, Pengyi

AU - Kamata, Seiichiro

PY - 2011/6

Y1 - 2011/6

N2 - Generally, two problems of bag-of-features in image retrieval are still considered unsolved: one is that spatial information about descriptors is not employed well, which affects the accuracy of retrieval; the other is that the trade-off between vocabulary size and good precision, which decides the storage and retrieval performance. In this paper, we propose a novel approach called Hilbert scan based bag-of-features (HS-BoF) for image retrieval. Firstly, Hilbert scan based tree representation (HSBT) is studied, which is built based on the local descriptors while spatial relationships are added into the nodes by a novel grouping rule, resulting of a tree structure for each image. Further, we give two ways of codebook production based on HSBT: multi-layer codebook and multi-size codebook. Owing to the properties of Hilbert scanning and the merits of our grouping method, sub-regions of the tree are not only flexible to the distribution of local patches but also have hierarchical relations. Extensive experiments on caltech-256, 13-scene and 1 million ImageNet images show that HS-BoF obtains higher accuracy with less memory usage.

AB - Generally, two problems of bag-of-features in image retrieval are still considered unsolved: one is that spatial information about descriptors is not employed well, which affects the accuracy of retrieval; the other is that the trade-off between vocabulary size and good precision, which decides the storage and retrieval performance. In this paper, we propose a novel approach called Hilbert scan based bag-of-features (HS-BoF) for image retrieval. Firstly, Hilbert scan based tree representation (HSBT) is studied, which is built based on the local descriptors while spatial relationships are added into the nodes by a novel grouping rule, resulting of a tree structure for each image. Further, we give two ways of codebook production based on HSBT: multi-layer codebook and multi-size codebook. Owing to the properties of Hilbert scanning and the merits of our grouping method, sub-regions of the tree are not only flexible to the distribution of local patches but also have hierarchical relations. Extensive experiments on caltech-256, 13-scene and 1 million ImageNet images show that HS-BoF obtains higher accuracy with less memory usage.

KW - Bag-of-features

KW - Feature representation

KW - Hilbert scanning

KW - Image search

UR - http://www.scopus.com/inward/record.url?scp=79957954955&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79957954955&partnerID=8YFLogxK

U2 - 10.1587/transinf.E94.D.1260

DO - 10.1587/transinf.E94.D.1260

M3 - Article

AN - SCOPUS:79957954955

VL - E94-D

SP - 1260

EP - 1268

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 6

ER -