Selection of Optimal Object Features in Object-Based Image Analysis Using Filter-Based Algorithms View Full Text


Ontology type: schema:ScholarlyArticle     


Article Info

DATE

2018-07-21

AUTHORS

Ismail Colkesen, Taskin Kavzoglu

ABSTRACT

With the increase in spatial resolution of recent sensors, object-based image analysis (OBIA) has gained importance for producing detailed land use maps. One of the main advantages of OBIA is that a variety of spectral, spatial and textural features can be extracted for the segmented image objects that are later utilized in classification. However, using a large number of features not only increases the required computational time, but also requires a large number of ground samples, which is unavailable in most cases. For these reasons, feature selection (FS) has become an important research topic for OBIA based classification studies. In this study, three filter-based FS algorithms namely, Chi square, information gain and ReliefF were applied to determine the most effective object features that ensure high separability among landscape features. For this purpose, importance degree (i.e. ranks) of 110 input object features were firstly estimated by the algorithms, and correlation-based merit function was then applied to determine optimum feature subset size. Multi-resolution segmentation algorithm was applied for segmenting a WorldView-2 image. Support vector machine, random forest and nearest neighbour classifiers were all utilized to classify segmented image objects using the selected object features. Results revealed that the FS algorithms were effective for selecting the most relevant features. Also, the classifiers produced the highest performances with 24 out of 110 features selected by the information gain (IG) algorithm. Particularly, the support vector machine classifier produced the highest overall accuracy (92.00%) with 24 selected features determined by the IG algorithm. A significant improvement of about 4% was achieved by applying FS procedures that was found statistically significant in terms of Wilcoxon signed-ranks test. More... »

PAGES

1233-1242

References to SciGraph publications

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/s12524-018-0807-x

DOI

http://dx.doi.org/10.1007/s12524-018-0807-x

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1105740143


Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
Incoming Citations Browse incoming citations for this publication using opencitations.net

JSON-LD is the canonical representation for SciGraph data.

TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

[
  {
    "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
    "about": [
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/08", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Information and Computing Sciences", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0801", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Artificial Intelligence and Image Processing", 
        "type": "DefinedTerm"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "Department of Geomatics Engineering, Gebze Technical University, 41400, Gebze, Kocaeli, Turkey", 
          "id": "http://www.grid.ac/institutes/grid.448834.7", 
          "name": [
            "Department of Geomatics Engineering, Gebze Technical University, 41400, Gebze, Kocaeli, Turkey"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Colkesen", 
        "givenName": "Ismail", 
        "id": "sg:person.013251461425.95", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013251461425.95"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Department of Geomatics Engineering, Gebze Technical University, 41400, Gebze, Kocaeli, Turkey", 
          "id": "http://www.grid.ac/institutes/grid.448834.7", 
          "name": [
            "Department of Geomatics Engineering, Gebze Technical University, 41400, Gebze, Kocaeli, Turkey"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Kavzoglu", 
        "givenName": "Taskin", 
        "id": "sg:person.010500456471.94", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.010500456471.94"
        ], 
        "type": "Person"
      }
    ], 
    "citation": [
      {
        "id": "sg:pub.10.1007/11552253_40", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1013464544", 
          "https://doi.org/10.1007/11552253_40"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1023/a:1010933404324", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1024739340", 
          "https://doi.org/10.1023/a:1010933404324"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1023/a:1025667309714", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1005679960", 
          "https://doi.org/10.1023/a:1025667309714"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/978-1-4757-2440-0", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1027312764", 
          "https://doi.org/10.1007/978-1-4757-2440-0"
        ], 
        "type": "CreativeWork"
      }
    ], 
    "datePublished": "2018-07-21", 
    "datePublishedReg": "2018-07-21", 
    "description": "With the increase in spatial resolution of recent sensors, object-based image analysis (OBIA) has gained importance for producing detailed land use maps. One of the main advantages of OBIA is that a variety of spectral, spatial and textural features can be extracted for the segmented image objects that are later utilized in classification. However, using a large number of features not only increases the required computational time, but also requires a large number of ground samples, which is unavailable in most cases. For these reasons, feature selection (FS) has become an important research topic for OBIA based classification studies. In this study, three filter-based FS algorithms namely, Chi square, information gain and ReliefF were applied to determine the most effective object features that ensure high separability among landscape features. For this purpose, importance degree (i.e. ranks) of 110 input object features were firstly estimated by the algorithms, and correlation-based merit function was then applied to determine optimum feature subset size. Multi-resolution segmentation algorithm was applied for segmenting a WorldView-2 image. Support vector machine, random forest and nearest neighbour classifiers were all utilized to classify segmented image objects using the selected object features. Results revealed that the FS algorithms were effective for selecting the most relevant features. Also, the classifiers produced the highest performances with 24 out of 110 features selected by the information gain (IG) algorithm. Particularly, the support vector machine classifier produced the highest overall accuracy (92.00%) with 24 selected features determined by the IG algorithm. A significant improvement of about 4% was achieved by applying FS procedures that was found statistically significant in terms of Wilcoxon signed-ranks test.", 
    "genre": "article", 
    "id": "sg:pub.10.1007/s12524-018-0807-x", 
    "inLanguage": "en", 
    "isAccessibleForFree": false, 
    "isPartOf": [
      {
        "id": "sg:journal.1136874", 
        "issn": [
          "0255-660X", 
          "0974-3006"
        ], 
        "name": "Journal of the Indian Society of Remote Sensing", 
        "publisher": "Springer Nature", 
        "type": "Periodical"
      }, 
      {
        "issueNumber": "8", 
        "type": "PublicationIssue"
      }, 
      {
        "type": "PublicationVolume", 
        "volumeNumber": "46"
      }
    ], 
    "keywords": [
      "object-based image analysis", 
      "segmented image objects", 
      "object features", 
      "FS algorithm", 
      "image objects", 
      "feature selection", 
      "multi-resolution segmentation algorithm", 
      "support vector machine classifier", 
      "information gain algorithm", 
      "feature subset size", 
      "support vector machine", 
      "vector machine classifier", 
      "nearest neighbour classifier", 
      "important research topic", 
      "image analysis", 
      "segmentation algorithm", 
      "vector machine", 
      "information gain", 
      "machine classifier", 
      "computational time", 
      "neighbour classifier", 
      "random forest", 
      "gain algorithm", 
      "IG algorithm", 
      "algorithm", 
      "relevant features", 
      "classifier", 
      "research topic", 
      "subset size", 
      "highest overall accuracy", 
      "high performance", 
      "textural features", 
      "recent sensors", 
      "WorldView-2 images", 
      "large number", 
      "objects", 
      "importance degree", 
      "main advantage", 
      "high separability", 
      "overall accuracy", 
      "classification studies", 
      "ReliefF", 
      "FS procedure", 
      "features", 
      "machine", 
      "images", 
      "merit function", 
      "classification", 
      "selection", 
      "sensors", 
      "significant improvement", 
      "accuracy", 
      "separability", 
      "performance", 
      "detailed land use maps", 
      "filter", 
      "maps", 
      "advantages", 
      "spatial resolution", 
      "topic", 
      "number", 
      "most cases", 
      "improvement", 
      "squares", 
      "terms", 
      "time", 
      "variety", 
      "gain", 
      "resolution", 
      "analysis", 
      "results", 
      "purpose", 
      "use maps", 
      "land use maps", 
      "reasons", 
      "function", 
      "procedure", 
      "size", 
      "importance", 
      "cases", 
      "forest", 
      "landscape features", 
      "degree", 
      "Wilcoxon signed-rank test", 
      "signed-rank test", 
      "test", 
      "study", 
      "chi square", 
      "increase", 
      "samples", 
      "ground samples", 
      "filter-based FS algorithms", 
      "effective object features", 
      "input object features", 
      "correlation-based merit function", 
      "optimum feature subset size", 
      "Optimal Object Features"
    ], 
    "name": "Selection of Optimal Object Features in Object-Based Image Analysis Using Filter-Based Algorithms", 
    "pagination": "1233-1242", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1105740143"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/s12524-018-0807-x"
        ]
      }
    ], 
    "sameAs": [
      "https://doi.org/10.1007/s12524-018-0807-x", 
      "https://app.dimensions.ai/details/publication/pub.1105740143"
    ], 
    "sdDataset": "articles", 
    "sdDatePublished": "2021-12-01T19:41", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20211201/entities/gbq_results/article/article_770.jsonl", 
    "type": "ScholarlyArticle", 
    "url": "https://doi.org/10.1007/s12524-018-0807-x"
  }
]
 

Download the RDF metadata as:  json-ld nt turtle xml License info

HOW TO GET THIS DATA PROGRAMMATICALLY:

JSON-LD is a popular format for linked data which is fully compatible with JSON.

curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/s12524-018-0807-x'

N-Triples is a line-based linked data format ideal for batch operations.

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/s12524-018-0807-x'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/s12524-018-0807-x'

RDF/XML is a standard XML format for linked data.

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/s12524-018-0807-x'


 

This table displays all metadata directly associated to this object as RDF triples.

178 TRIPLES      22 PREDICATES      126 URIs      114 LITERALS      6 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/s12524-018-0807-x schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author N98b45529e3a54fe1ba0cccc320c77efa
4 schema:citation sg:pub.10.1007/11552253_40
5 sg:pub.10.1007/978-1-4757-2440-0
6 sg:pub.10.1023/a:1010933404324
7 sg:pub.10.1023/a:1025667309714
8 schema:datePublished 2018-07-21
9 schema:datePublishedReg 2018-07-21
10 schema:description With the increase in spatial resolution of recent sensors, object-based image analysis (OBIA) has gained importance for producing detailed land use maps. One of the main advantages of OBIA is that a variety of spectral, spatial and textural features can be extracted for the segmented image objects that are later utilized in classification. However, using a large number of features not only increases the required computational time, but also requires a large number of ground samples, which is unavailable in most cases. For these reasons, feature selection (FS) has become an important research topic for OBIA based classification studies. In this study, three filter-based FS algorithms namely, Chi square, information gain and ReliefF were applied to determine the most effective object features that ensure high separability among landscape features. For this purpose, importance degree (i.e. ranks) of 110 input object features were firstly estimated by the algorithms, and correlation-based merit function was then applied to determine optimum feature subset size. Multi-resolution segmentation algorithm was applied for segmenting a WorldView-2 image. Support vector machine, random forest and nearest neighbour classifiers were all utilized to classify segmented image objects using the selected object features. Results revealed that the FS algorithms were effective for selecting the most relevant features. Also, the classifiers produced the highest performances with 24 out of 110 features selected by the information gain (IG) algorithm. Particularly, the support vector machine classifier produced the highest overall accuracy (92.00%) with 24 selected features determined by the IG algorithm. A significant improvement of about 4% was achieved by applying FS procedures that was found statistically significant in terms of Wilcoxon signed-ranks test.
11 schema:genre article
12 schema:inLanguage en
13 schema:isAccessibleForFree false
14 schema:isPartOf N4e498cecf42945e69ed3c03d723edc0b
15 N7a1447f974084276a554ba0245ba34e2
16 sg:journal.1136874
17 schema:keywords FS algorithm
18 FS procedure
19 IG algorithm
20 Optimal Object Features
21 ReliefF
22 Wilcoxon signed-rank test
23 WorldView-2 images
24 accuracy
25 advantages
26 algorithm
27 analysis
28 cases
29 chi square
30 classification
31 classification studies
32 classifier
33 computational time
34 correlation-based merit function
35 degree
36 detailed land use maps
37 effective object features
38 feature selection
39 feature subset size
40 features
41 filter
42 filter-based FS algorithms
43 forest
44 function
45 gain
46 gain algorithm
47 ground samples
48 high performance
49 high separability
50 highest overall accuracy
51 image analysis
52 image objects
53 images
54 importance
55 importance degree
56 important research topic
57 improvement
58 increase
59 information gain
60 information gain algorithm
61 input object features
62 land use maps
63 landscape features
64 large number
65 machine
66 machine classifier
67 main advantage
68 maps
69 merit function
70 most cases
71 multi-resolution segmentation algorithm
72 nearest neighbour classifier
73 neighbour classifier
74 number
75 object features
76 object-based image analysis
77 objects
78 optimum feature subset size
79 overall accuracy
80 performance
81 procedure
82 purpose
83 random forest
84 reasons
85 recent sensors
86 relevant features
87 research topic
88 resolution
89 results
90 samples
91 segmentation algorithm
92 segmented image objects
93 selection
94 sensors
95 separability
96 signed-rank test
97 significant improvement
98 size
99 spatial resolution
100 squares
101 study
102 subset size
103 support vector machine
104 support vector machine classifier
105 terms
106 test
107 textural features
108 time
109 topic
110 use maps
111 variety
112 vector machine
113 vector machine classifier
114 schema:name Selection of Optimal Object Features in Object-Based Image Analysis Using Filter-Based Algorithms
115 schema:pagination 1233-1242
116 schema:productId N9d31237780d149c8ba19a40db1b6841e
117 Na6e90e057a67485281cae24eedd0ddeb
118 schema:sameAs https://app.dimensions.ai/details/publication/pub.1105740143
119 https://doi.org/10.1007/s12524-018-0807-x
120 schema:sdDatePublished 2021-12-01T19:41
121 schema:sdLicense https://scigraph.springernature.com/explorer/license/
122 schema:sdPublisher Nbf36b43ca3984444bed3d23520429120
123 schema:url https://doi.org/10.1007/s12524-018-0807-x
124 sgo:license sg:explorer/license/
125 sgo:sdDataset articles
126 rdf:type schema:ScholarlyArticle
127 N4e498cecf42945e69ed3c03d723edc0b schema:issueNumber 8
128 rdf:type schema:PublicationIssue
129 N7a1447f974084276a554ba0245ba34e2 schema:volumeNumber 46
130 rdf:type schema:PublicationVolume
131 N8748fe48be3941518941db97e5feca24 rdf:first sg:person.010500456471.94
132 rdf:rest rdf:nil
133 N98b45529e3a54fe1ba0cccc320c77efa rdf:first sg:person.013251461425.95
134 rdf:rest N8748fe48be3941518941db97e5feca24
135 N9d31237780d149c8ba19a40db1b6841e schema:name dimensions_id
136 schema:value pub.1105740143
137 rdf:type schema:PropertyValue
138 Na6e90e057a67485281cae24eedd0ddeb schema:name doi
139 schema:value 10.1007/s12524-018-0807-x
140 rdf:type schema:PropertyValue
141 Nbf36b43ca3984444bed3d23520429120 schema:name Springer Nature - SN SciGraph project
142 rdf:type schema:Organization
143 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
144 schema:name Information and Computing Sciences
145 rdf:type schema:DefinedTerm
146 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
147 schema:name Artificial Intelligence and Image Processing
148 rdf:type schema:DefinedTerm
149 sg:journal.1136874 schema:issn 0255-660X
150 0974-3006
151 schema:name Journal of the Indian Society of Remote Sensing
152 schema:publisher Springer Nature
153 rdf:type schema:Periodical
154 sg:person.010500456471.94 schema:affiliation grid-institutes:grid.448834.7
155 schema:familyName Kavzoglu
156 schema:givenName Taskin
157 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.010500456471.94
158 rdf:type schema:Person
159 sg:person.013251461425.95 schema:affiliation grid-institutes:grid.448834.7
160 schema:familyName Colkesen
161 schema:givenName Ismail
162 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013251461425.95
163 rdf:type schema:Person
164 sg:pub.10.1007/11552253_40 schema:sameAs https://app.dimensions.ai/details/publication/pub.1013464544
165 https://doi.org/10.1007/11552253_40
166 rdf:type schema:CreativeWork
167 sg:pub.10.1007/978-1-4757-2440-0 schema:sameAs https://app.dimensions.ai/details/publication/pub.1027312764
168 https://doi.org/10.1007/978-1-4757-2440-0
169 rdf:type schema:CreativeWork
170 sg:pub.10.1023/a:1010933404324 schema:sameAs https://app.dimensions.ai/details/publication/pub.1024739340
171 https://doi.org/10.1023/a:1010933404324
172 rdf:type schema:CreativeWork
173 sg:pub.10.1023/a:1025667309714 schema:sameAs https://app.dimensions.ai/details/publication/pub.1005679960
174 https://doi.org/10.1023/a:1025667309714
175 rdf:type schema:CreativeWork
176 grid-institutes:grid.448834.7 schema:alternateName Department of Geomatics Engineering, Gebze Technical University, 41400, Gebze, Kocaeli, Turkey
177 schema:name Department of Geomatics Engineering, Gebze Technical University, 41400, Gebze, Kocaeli, Turkey
178 rdf:type schema:Organization
 




Preview window. Press ESC to close (or click here)


...