STO Project

Introduction and Objectives

The need of classifying remotely sensed images and characterise the image informative contents in terms of objects and their characteristics is becoming an urgent problem to solve for Earth Observation discipline, especially with the support of automated tools. The more the resolution of satellite sensors increases and the more the level of detail becomes higher and higher; not only the identification and characterization of the smallest possible object that can be detected in one image is important, but also the possibility to correctly group together small objects to create larger and more complex objects is of paramount importance.
The Support to TOpology (STO) TRP project aims at analysing the latter aspect, identifying from users the needs and the methodologies to group together simple objects, and analysing them with experts in the topology domain to verify how the user requests can be satisfied with the theoretical topological approach. Moreover, a prototype software tool is implemented to provide the user with an instrument to develop topological rules to create complex objects; the software tool will be tested over a set of few use cases that are identified and run by the end users to verify the proposed approach. Finally the project aims at proposing, to the geographic objects description standardization community (e.g. the CityGML development community), the project findings, in terms of objects description and classification means.

The main project objectives are:

  • The identification of user's needs for the aggregation of simple objects to create complex objects, and the identification of the topological operators / rules to satisfy this process
  • The formalization of the topological tools into a topological approach to provide indication about how the existing GIS / EO standards are (not) adequate to describe the complex objects
  • The inclusion also of ontology / semantics, in order to permit to define and identify objects and spatial relations in terms familiar to the related applications
  • The implementation of a software prototype to define and apply topological rules for complex objects identification
  • The definition of test cases directly operated by the end users, to verify the effectiveness of the implemented approach
  • The improvement of GIS / EO existing standards to take into consideration also complex objects.

The completion of the project objectives shall further increase the usefulness of EO images for automated detection of complex structures (e.g.: a nuclear plant or a chemical factory), as the result of the combination of different objects or of groups of objects. The detection of these objects and their handling within a GIS environment shall permit to opens up new opportunities for EO products and services.

Expected Results

The main project expectations include:

  • The verification that a topological approach can be applied to EO products for the identification and classification of complex objects and therefore the formalisation and validation of a set of rules to identify complex objects over a wide area from very high resolution EO images
  • The creation of software tool/platform in support to more automated identification of complex objects or features in EO images
  • The verification/demonstration of the usability of these topological relationships / rules and developed tool with practical use cases such as land use classification
  • The proposal to the Eo standardization community of the project findings, in terms of, for instance, the need of modifying existing standards to account for different level of details /aggregations



Contributors to this page: Michele Iapaolo
Stefano natali .

Page last modified on Friday 29 of April 2016 12:43:47 CEST by Michele Iapaolo.