Filters, Random Fields and Maximum Entropy (FRAME): Towards a Unified Theory for Texture Modeling View Full Text


Ontology type: schema:ScholarlyArticle      Open Access: True


Article Info

DATE

1998-03

AUTHORS

Song Chun Zhu, Yingnian Wu, David Mumford

ABSTRACT

This article presents a statistical theory for texture modeling. This theory combines filtering theory and Markov random field modeling through the maximum entropy principle, and interprets and clarifies many previous concepts and methods for texture analysis and synthesis from a unified point of view. Our theory characterizes the ensemble of images I with the same texture appearance by a probability distribution f(I) on a random field, and the objective of texture modeling is to make inference about f(I), given a set of observed texture examples.In our theory, texture modeling consists of two steps. (1) A set of filters is selected from a general filter bank to capture features of the texture, these filters are applied to observed texture images, and the histograms of the filtered images are extracted. These histograms are estimates of the marginal distributions of f( I). This step is called feature extraction. (2) The maximum entropy principle is employed to derive a distribution p(I), which is restricted to have the same marginal distributions as those in (1). This p(I) is considered as an estimate of f( I). This step is called feature fusion. A stepwise algorithm is proposed to choose filters from a general filter bank. The resulting model, called FRAME (Filters, Random fields And Maximum Entropy), is a Markov random field (MRF) model, but with a much enriched vocabulary and hence much stronger descriptive ability than the previous MRF models used for texture modeling. Gibbs sampler is adopted to synthesize texture images by drawing typical samples from p(I), thus the model is verified by seeing whether the synthesized texture images have similar visual appearances to the texture images being modeled. Experiments on a variety of 1D and 2D textures are described to illustrate our theory and to show the performance of our algorithms. These experiments demonstrate that many textures which are previously considered as from different categories can be modeled and synthesized in a common framework. More... »

PAGES

107-126

References to SciGraph publications

Identifiers

URI

http://scigraph.springernature.com/pub.10.1023/a:1007925832420

DOI

http://dx.doi.org/10.1023/a:1007925832420

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1005653889


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 Computer Science, Stanford University, 94305, Stanford, CA", 
          "id": "http://www.grid.ac/institutes/grid.168010.e", 
          "name": [
            "Department of Computer Science, Stanford University, 94305, Stanford, CA"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Zhu", 
        "givenName": "Song Chun", 
        "id": "sg:person.012430434515.36", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012430434515.36"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Department of Statistics, University of Michigan, 48109, Ann Arbor, MI", 
          "id": "http://www.grid.ac/institutes/grid.214458.e", 
          "name": [
            "Department of Statistics, University of Michigan, 48109, Ann Arbor, MI"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Wu", 
        "givenName": "Yingnian", 
        "id": "sg:person.0760051514.84", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0760051514.84"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Division of Applied Math, Brown University, 02912, Providence, RI", 
          "id": "http://www.grid.ac/institutes/grid.40263.33", 
          "name": [
            "Division of Applied Math, Brown University, 02912, Providence, RI"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Mumford", 
        "givenName": "David", 
        "id": "sg:person.01243511632.19", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01243511632.19"
        ], 
        "type": "Person"
      }
    ], 
    "citation": [
      {
        "id": "sg:pub.10.1007/978-3-642-97522-6", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1053595130", 
          "https://doi.org/10.1007/978-3-642-97522-6"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/bf00978978", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1037053331", 
          "https://doi.org/10.1007/bf00978978"
        ], 
        "type": "CreativeWork"
      }
    ], 
    "datePublished": "1998-03", 
    "datePublishedReg": "1998-03-01", 
    "description": "This article presents a statistical theory for texture modeling. This theory combines filtering theory and Markov random field modeling through the maximum entropy principle, and interprets and clarifies many previous concepts and methods for texture analysis and synthesis from a unified point of view. Our theory characterizes the ensemble of images I with the same texture appearance by a probability distribution f(I) on a random field, and the objective of texture modeling is to make inference about f(I), given a set of observed texture examples.In our theory, texture modeling consists of two steps. (1) A set of filters is selected from a general filter bank to capture features of the texture, these filters are applied to observed texture images, and the histograms of the filtered images are extracted. These histograms are estimates of the marginal distributions of f( I). This step is called feature extraction. (2) The maximum entropy principle is employed to derive a distribution p(I), which is restricted to have the same marginal distributions as those in (1). This p(I) is considered as an estimate of f( I). This step is called feature fusion. A stepwise algorithm is proposed to choose filters from a general filter bank. The resulting model, called FRAME (Filters, Random fields And Maximum Entropy), is a Markov random field (MRF) model, but with a much enriched vocabulary and hence much stronger descriptive ability than the previous MRF models used for texture modeling. Gibbs sampler is adopted to synthesize texture images by drawing typical samples from p(I), thus the model is verified by seeing whether the synthesized texture images have similar visual appearances to the texture images being modeled. Experiments on a variety of 1D and 2D textures are described to illustrate our theory and to show the performance of our algorithms. These experiments demonstrate that many textures which are previously considered as from different categories can be modeled and synthesized in a common framework.", 
    "genre": "article", 
    "id": "sg:pub.10.1023/a:1007925832420", 
    "isAccessibleForFree": true, 
    "isPartOf": [
      {
        "id": "sg:journal.1032807", 
        "issn": [
          "0920-5691", 
          "1573-1405"
        ], 
        "name": "International Journal of Computer Vision", 
        "publisher": "Springer Nature", 
        "type": "Periodical"
      }, 
      {
        "issueNumber": "2", 
        "type": "PublicationIssue"
      }, 
      {
        "type": "PublicationVolume", 
        "volumeNumber": "27"
      }
    ], 
    "keywords": [
      "maximum entropy principle", 
      "random fields", 
      "entropy principle", 
      "general filter banks", 
      "marginal distributions", 
      "texture modeling", 
      "Markov random field model", 
      "same marginal distributions", 
      "random field model", 
      "Markov random field", 
      "statistical theory", 
      "Gibbs sampler", 
      "stronger descriptive ability", 
      "probability distribution", 
      "maximum entropy", 
      "unified point", 
      "filter bank", 
      "MRF model", 
      "field model", 
      "modeling consists", 
      "unified theory", 
      "theory", 
      "descriptive ability", 
      "stepwise algorithm", 
      "variety of 1D", 
      "modeling", 
      "set of filters", 
      "algorithm", 
      "synthesized texture image", 
      "filter", 
      "field", 
      "common framework", 
      "model", 
      "texture images", 
      "distribution", 
      "estimates", 
      "entropy", 
      "set", 
      "typical samples", 
      "inference", 
      "ensemble", 
      "principles", 
      "image I", 
      "histogram", 
      "previous concepts", 
      "step", 
      "similar visual appearance", 
      "texture analysis", 
      "sampler", 
      "point", 
      "framework", 
      "experiments", 
      "texture appearance", 
      "performance", 
      "frame", 
      "concept", 
      "clarifies", 
      "consist", 
      "images", 
      "texture", 
      "different categories", 
      "features", 
      "analysis", 
      "feature extraction", 
      "variety", 
      "appearance", 
      "banks", 
      "objective", 
      "view", 
      "feature fusion", 
      "visual appearance", 
      "article", 
      "interpret", 
      "categories", 
      "samples", 
      "ability", 
      "fusion", 
      "extraction", 
      "synthesis", 
      "vocabulary", 
      "example", 
      "method"
    ], 
    "name": "Filters, Random Fields and Maximum Entropy (FRAME): Towards a Unified Theory for Texture Modeling", 
    "pagination": "107-126", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1005653889"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1023/a:1007925832420"
        ]
      }
    ], 
    "sameAs": [
      "https://doi.org/10.1023/a:1007925832420", 
      "https://app.dimensions.ai/details/publication/pub.1005653889"
    ], 
    "sdDataset": "articles", 
    "sdDatePublished": "2022-09-02T15:48", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20220902/entities/gbq_results/article/article_296.jsonl", 
    "type": "ScholarlyArticle", 
    "url": "https://doi.org/10.1023/a:1007925832420"
  }
]
 

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.10.1023/a:1007925832420'

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

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1023/a:1007925832420'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1023/a:1007925832420'

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

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1023/a:1007925832420'


 

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

167 TRIPLES      21 PREDICATES      108 URIs      98 LITERALS      6 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1023/a:1007925832420 schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author N99930a9f50cd4903ace35171b98bc102
4 schema:citation sg:pub.10.1007/978-3-642-97522-6
5 sg:pub.10.1007/bf00978978
6 schema:datePublished 1998-03
7 schema:datePublishedReg 1998-03-01
8 schema:description This article presents a statistical theory for texture modeling. This theory combines filtering theory and Markov random field modeling through the maximum entropy principle, and interprets and clarifies many previous concepts and methods for texture analysis and synthesis from a unified point of view. Our theory characterizes the ensemble of images I with the same texture appearance by a probability distribution f(I) on a random field, and the objective of texture modeling is to make inference about f(I), given a set of observed texture examples.In our theory, texture modeling consists of two steps. (1) A set of filters is selected from a general filter bank to capture features of the texture, these filters are applied to observed texture images, and the histograms of the filtered images are extracted. These histograms are estimates of the marginal distributions of f( I). This step is called feature extraction. (2) The maximum entropy principle is employed to derive a distribution p(I), which is restricted to have the same marginal distributions as those in (1). This p(I) is considered as an estimate of f( I). This step is called feature fusion. A stepwise algorithm is proposed to choose filters from a general filter bank. The resulting model, called FRAME (Filters, Random fields And Maximum Entropy), is a Markov random field (MRF) model, but with a much enriched vocabulary and hence much stronger descriptive ability than the previous MRF models used for texture modeling. Gibbs sampler is adopted to synthesize texture images by drawing typical samples from p(I), thus the model is verified by seeing whether the synthesized texture images have similar visual appearances to the texture images being modeled. Experiments on a variety of 1D and 2D textures are described to illustrate our theory and to show the performance of our algorithms. These experiments demonstrate that many textures which are previously considered as from different categories can be modeled and synthesized in a common framework.
9 schema:genre article
10 schema:isAccessibleForFree true
11 schema:isPartOf N2e563f3435e34de3a1e4333962b394bc
12 N90ae063df4be4f548582bb56645a6a2e
13 sg:journal.1032807
14 schema:keywords Gibbs sampler
15 MRF model
16 Markov random field
17 Markov random field model
18 ability
19 algorithm
20 analysis
21 appearance
22 article
23 banks
24 categories
25 clarifies
26 common framework
27 concept
28 consist
29 descriptive ability
30 different categories
31 distribution
32 ensemble
33 entropy
34 entropy principle
35 estimates
36 example
37 experiments
38 extraction
39 feature extraction
40 feature fusion
41 features
42 field
43 field model
44 filter
45 filter bank
46 frame
47 framework
48 fusion
49 general filter banks
50 histogram
51 image I
52 images
53 inference
54 interpret
55 marginal distributions
56 maximum entropy
57 maximum entropy principle
58 method
59 model
60 modeling
61 modeling consists
62 objective
63 performance
64 point
65 previous concepts
66 principles
67 probability distribution
68 random field model
69 random fields
70 same marginal distributions
71 sampler
72 samples
73 set
74 set of filters
75 similar visual appearance
76 statistical theory
77 step
78 stepwise algorithm
79 stronger descriptive ability
80 synthesis
81 synthesized texture image
82 texture
83 texture analysis
84 texture appearance
85 texture images
86 texture modeling
87 theory
88 typical samples
89 unified point
90 unified theory
91 variety
92 variety of 1D
93 view
94 visual appearance
95 vocabulary
96 schema:name Filters, Random Fields and Maximum Entropy (FRAME): Towards a Unified Theory for Texture Modeling
97 schema:pagination 107-126
98 schema:productId N454a96a832114b7fb2ccb5542e7169e2
99 Ne547a75996b94f8b8905435f0fc2d428
100 schema:sameAs https://app.dimensions.ai/details/publication/pub.1005653889
101 https://doi.org/10.1023/a:1007925832420
102 schema:sdDatePublished 2022-09-02T15:48
103 schema:sdLicense https://scigraph.springernature.com/explorer/license/
104 schema:sdPublisher N80e70493787a44eb924a2b0e35b29a8c
105 schema:url https://doi.org/10.1023/a:1007925832420
106 sgo:license sg:explorer/license/
107 sgo:sdDataset articles
108 rdf:type schema:ScholarlyArticle
109 N2e563f3435e34de3a1e4333962b394bc schema:volumeNumber 27
110 rdf:type schema:PublicationVolume
111 N454a96a832114b7fb2ccb5542e7169e2 schema:name dimensions_id
112 schema:value pub.1005653889
113 rdf:type schema:PropertyValue
114 N80e70493787a44eb924a2b0e35b29a8c schema:name Springer Nature - SN SciGraph project
115 rdf:type schema:Organization
116 N90ae063df4be4f548582bb56645a6a2e schema:issueNumber 2
117 rdf:type schema:PublicationIssue
118 N99930a9f50cd4903ace35171b98bc102 rdf:first sg:person.012430434515.36
119 rdf:rest Na3b565e39b444640a7f50247774d8254
120 Na3b565e39b444640a7f50247774d8254 rdf:first sg:person.0760051514.84
121 rdf:rest Nc99e80b90dac45248971bdcf5e97ac07
122 Nc99e80b90dac45248971bdcf5e97ac07 rdf:first sg:person.01243511632.19
123 rdf:rest rdf:nil
124 Ne547a75996b94f8b8905435f0fc2d428 schema:name doi
125 schema:value 10.1023/a:1007925832420
126 rdf:type schema:PropertyValue
127 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
128 schema:name Information and Computing Sciences
129 rdf:type schema:DefinedTerm
130 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
131 schema:name Artificial Intelligence and Image Processing
132 rdf:type schema:DefinedTerm
133 sg:journal.1032807 schema:issn 0920-5691
134 1573-1405
135 schema:name International Journal of Computer Vision
136 schema:publisher Springer Nature
137 rdf:type schema:Periodical
138 sg:person.012430434515.36 schema:affiliation grid-institutes:grid.168010.e
139 schema:familyName Zhu
140 schema:givenName Song Chun
141 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012430434515.36
142 rdf:type schema:Person
143 sg:person.01243511632.19 schema:affiliation grid-institutes:grid.40263.33
144 schema:familyName Mumford
145 schema:givenName David
146 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01243511632.19
147 rdf:type schema:Person
148 sg:person.0760051514.84 schema:affiliation grid-institutes:grid.214458.e
149 schema:familyName Wu
150 schema:givenName Yingnian
151 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0760051514.84
152 rdf:type schema:Person
153 sg:pub.10.1007/978-3-642-97522-6 schema:sameAs https://app.dimensions.ai/details/publication/pub.1053595130
154 https://doi.org/10.1007/978-3-642-97522-6
155 rdf:type schema:CreativeWork
156 sg:pub.10.1007/bf00978978 schema:sameAs https://app.dimensions.ai/details/publication/pub.1037053331
157 https://doi.org/10.1007/bf00978978
158 rdf:type schema:CreativeWork
159 grid-institutes:grid.168010.e schema:alternateName Department of Computer Science, Stanford University, 94305, Stanford, CA
160 schema:name Department of Computer Science, Stanford University, 94305, Stanford, CA
161 rdf:type schema:Organization
162 grid-institutes:grid.214458.e schema:alternateName Department of Statistics, University of Michigan, 48109, Ann Arbor, MI
163 schema:name Department of Statistics, University of Michigan, 48109, Ann Arbor, MI
164 rdf:type schema:Organization
165 grid-institutes:grid.40263.33 schema:alternateName Division of Applied Math, Brown University, 02912, Providence, RI
166 schema:name Division of Applied Math, Brown University, 02912, Providence, RI
167 rdf:type schema:Organization
 




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


...