A Novel Discriminating and Relative Global Spatial Image Representation with Applications in CBIR
The requirement for effective image search, which motivates the use of Content-Based Image Retrieval (CBIR) and the search of similar multimedia contents on the basis of user query, remains an open research problem for computer vision applications. The application domains for Bag of Visual Words (Bo...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2018-11-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/8/11/2242 |
id |
doaj-38d29c847a5a41f4b0dd4e6b9b7ba01f |
---|---|
record_format |
Article |
spelling |
doaj-38d29c847a5a41f4b0dd4e6b9b7ba01f2020-11-25T01:32:41ZengMDPI AGApplied Sciences2076-34172018-11-01811224210.3390/app8112242app8112242A Novel Discriminating and Relative Global Spatial Image Representation with Applications in CBIRBushra Zafar0Rehan Ashraf1Nouman Ali2Muhammad Kashif Iqbal3Muhammad Sajid4Saadat Hanif Dar5Naeem Iqbal Ratyal6Department of Computer Science, National Textile University, Faisalabad 38000, PakistanDepartment of Computer Science, National Textile University, Faisalabad 38000, PakistanDepartment of Software Engineering, Mirpur University of Science & Technology, Mirpur AJK 10250, PakistanDepartment of Mathematics, Government College University, Faisalabad 38000, PakistanDepartment of Electrical Engineering, Mirpur University of Science & Technology, Mirpur AJK 10250, PakistanDepartment of Software Engineering, Mirpur University of Science & Technology, Mirpur AJK 10250, PakistanDepartment of Electrical Engineering, Mirpur University of Science & Technology, Mirpur AJK 10250, PakistanThe requirement for effective image search, which motivates the use of Content-Based Image Retrieval (CBIR) and the search of similar multimedia contents on the basis of user query, remains an open research problem for computer vision applications. The application domains for Bag of Visual Words (BoVW) based image representations are object recognition, image classification and content-based image analysis. Interest point detectors are quantized in the feature space and the final histogram or image signature do not retain any detail about co-occurrences of features in the 2D image space. This spatial information is crucial, as it adversely affects the performance of an image classification-based model. The most notable contribution in this context is Spatial Pyramid Matching (SPM), which captures the absolute spatial distribution of visual words. However, SPM is sensitive to image transformations such as rotation, flipping and translation. When images are not well-aligned, SPM may lose its discriminative power. This paper introduces a novel approach to encoding the relative spatial information for histogram-based representation of the BoVW model. This is established by computing the global geometric relationship between pairs of identical visual words with respect to the centroid of an image. The proposed research is evaluated by using five different datasets. Comprehensive experiments demonstrate the robustness of the proposed image representation as compared to the state-of-the-art methods in terms of precision and recall values.https://www.mdpi.com/2076-3417/8/11/2242image analysisimage retrievalspatial informationimage classificationcomputer vision |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Bushra Zafar Rehan Ashraf Nouman Ali Muhammad Kashif Iqbal Muhammad Sajid Saadat Hanif Dar Naeem Iqbal Ratyal |
spellingShingle |
Bushra Zafar Rehan Ashraf Nouman Ali Muhammad Kashif Iqbal Muhammad Sajid Saadat Hanif Dar Naeem Iqbal Ratyal A Novel Discriminating and Relative Global Spatial Image Representation with Applications in CBIR Applied Sciences image analysis image retrieval spatial information image classification computer vision |
author_facet |
Bushra Zafar Rehan Ashraf Nouman Ali Muhammad Kashif Iqbal Muhammad Sajid Saadat Hanif Dar Naeem Iqbal Ratyal |
author_sort |
Bushra Zafar |
title |
A Novel Discriminating and Relative Global Spatial Image Representation with Applications in CBIR |
title_short |
A Novel Discriminating and Relative Global Spatial Image Representation with Applications in CBIR |
title_full |
A Novel Discriminating and Relative Global Spatial Image Representation with Applications in CBIR |
title_fullStr |
A Novel Discriminating and Relative Global Spatial Image Representation with Applications in CBIR |
title_full_unstemmed |
A Novel Discriminating and Relative Global Spatial Image Representation with Applications in CBIR |
title_sort |
novel discriminating and relative global spatial image representation with applications in cbir |
publisher |
MDPI AG |
series |
Applied Sciences |
issn |
2076-3417 |
publishDate |
2018-11-01 |
description |
The requirement for effective image search, which motivates the use of Content-Based Image Retrieval (CBIR) and the search of similar multimedia contents on the basis of user query, remains an open research problem for computer vision applications. The application domains for Bag of Visual Words (BoVW) based image representations are object recognition, image classification and content-based image analysis. Interest point detectors are quantized in the feature space and the final histogram or image signature do not retain any detail about co-occurrences of features in the 2D image space. This spatial information is crucial, as it adversely affects the performance of an image classification-based model. The most notable contribution in this context is Spatial Pyramid Matching (SPM), which captures the absolute spatial distribution of visual words. However, SPM is sensitive to image transformations such as rotation, flipping and translation. When images are not well-aligned, SPM may lose its discriminative power. This paper introduces a novel approach to encoding the relative spatial information for histogram-based representation of the BoVW model. This is established by computing the global geometric relationship between pairs of identical visual words with respect to the centroid of an image. The proposed research is evaluated by using five different datasets. Comprehensive experiments demonstrate the robustness of the proposed image representation as compared to the state-of-the-art methods in terms of precision and recall values. |
topic |
image analysis image retrieval spatial information image classification computer vision |
url |
https://www.mdpi.com/2076-3417/8/11/2242 |
work_keys_str_mv |
AT bushrazafar anoveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT rehanashraf anoveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT noumanali anoveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT muhammadkashifiqbal anoveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT muhammadsajid anoveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT saadathanifdar anoveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT naeemiqbalratyal anoveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT bushrazafar noveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT rehanashraf noveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT noumanali noveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT muhammadkashifiqbal noveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT muhammadsajid noveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT saadathanifdar noveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir AT naeemiqbalratyal noveldiscriminatingandrelativeglobalspatialimagerepresentationwithapplicationsincbir |
_version_ |
1725080546741059584 |