REMOTE SENSING IMAGE MINING USING ONTOLOGIES

1
REMOTE SENSING IMAGE MINING USING ONTOLOGIES

• Marcelino Pereira
• mpss_at_dpi.inpe.br
• Gilberto Câmara
• gilberto_at_dpi.inpe.br
• São José dos Campos, November 30, 2004.

2
Introduction

• Motivation
• Explosive growth of remote sensing datasets
• Strategic and fast information are demanded
• General objective
• Allow image mining in large repositories
• Through smart computational tools
• Providing for the users resources to get high
  level information from images using ontologies

3
Image Mining

• Data mining searches
• Valid patterns
• Previously unknown patterns
• Potentially useful patterns
• Understandable patterns
• Image mining extracts
• Strategic information
• Relationships and patterns
• Landscape aspects
• Challenges (image mining)
• Relative values
• Spatial information
• Multiple interpretation
• Patterns representation

Zhang et al., 2002
4
Ontologies

• Ontology
• Content theories, which have a general set of
  facts to be shared
• The main contribution is to identify specific
  classes of objects and relationships in a
  specific domain
• Specification mechanism
• Interoperability
• Reuse
• Clarity
• Coherence
• Extensibility

5
Image Ontology
Câmara et al., 2001

• Physical Ontology
• Describes the physical process of image
  generation
• Structural Ontology
• Concerns geometric, functional and descriptive
  structures that can be extracted

• Method Ontology
• Transformation algorithms
• physical level ?structural level
• Application Ontology
• Describes the vocabulary related to a generic
  domain
• Task Ontology
• Specializations of the App. Ontology

6
Image Ontology

• Example
• Application Ontology
• Forest and non-forest areas
• Task Ontology
• Cattle ranches
• Small farms
• Physical Ontology
• Statistical and morphological properties of the
  image
• Structural Ontology
• Region structure regular, fishbone, corridor and
  so on
• Method Ontology
• Algorithms and data structures to extract regions

Eymar Lopes, INPE
7
Image Ontology
Câmara et al., 2001

• Semantic Mediator
• Relates Image Ontology to Application Ontology
• Identify specific algorithms to extract the
  desired structures
• Maps concepts from domain ontology to extracted
  structures

• Mapping between an instance of a concept on the
  application domain and an instance of a concept
  on the structure domain
• Matching
• Set of matchings in a temporal instance
• Spatial configuration

8
Proposal 1st Phase(for a specific
deforestation pattern)
Building the Structural Ontology
Graph Generation
Graphs and Metrics
Prototypical Images
Segmentação e Rotulação
Segmentation and Labeling
Segmented Images
Application
Semantic Mediation Extraction of the
representative graphs (specialist)
Application Ontology Generation
Application Ontology
Spatial Patterns Typology
Building the Application Ontology
Graphs of the Pattern
9
Proposal 2nd Phase(for a specific
deforestation pattern)
Building the Structural Ontology
Graph Generation
Graphs and Metrics
Segmentação e Rotulação
Images
Segmentation and Labeling
Segmented Images
Graphs of the Pattern
Spatial Configurations
10
Application Domain

• Land Use and Cover Change
• Land use purpose to which its employed
  (agriculture, ranching)
• Land cover Physical status of its surface
  (forest, water)
• Changes bring environment, social and economics
  impacts
• The Amazon case complexity, dimensions and
  demands involved
• Deforestation 10.000.000 ha (1970s) ?
  59.000.000 ha (2000)
• Soil degradation, social conflicts, precarious
  urbanization
• Faster (and precise) the identification of such
  tendencies, higher the chances of preventing,
  managing and reducing their consequences

11
Object Representation Formalism

• Graphs
• Mathematical abstraction employed in many
  problems
• Well known and researched formalism
• Represents objects and relationships in an
  natural way
• Spation Configurations may be approached using
  graphs
• Graph inexact isomorphism

12
Building the Structural Ontology

• Extraction of areas (objects)
• Metrics generation
• perimeter area ratio
• fractal index (and so on)
• Graph mapping (objects, relationships, metrics)

Graph Generation
Graphs and Metrics
Segmentação e Rotulação
Segmentation and Labeling
Segmented Images
Images
Graphs and Metrics
13
Building the Structural Ontology
Application
Application Ontology Generation
Application Ontology
Spatial Patterns Typology
Diffuse, Bidirectional, Fishbone patterns
Mertens Lambin, 1997 Escada, 2003
14
Final Comments

• Challenges
• Image domain complexity
• Dimension and demands of Amazon case
• Resources and techniques from different areas
• Technologies development and integration
• Transforming objects in semantic entities
• Relevant scientific topic not yet solved
• INPE
• Has a rich remote sensing dataset
• Knows and research Amazon history
• Experience in image processing and software
  development

15
Thank you!