Interactions segmentation-classification dans un cadre multi-paradigme pour l'analyse d'images de télédétection

Translated title of the contribution: Study of segmentation-classification interactions within a multi-paradigm framework for remote sensing image analysis

Andrés Troya-Galvis, Pierre Gançarski, Laure Berti-Equille

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Segmentation and classification tasks are closely related in the remote sensing image analysis domain. Collaborative approaches allow interactions between segmentation and classification techniques in order to mutually improve both results. In this article we present a generic collaborative framework for segmentation and classification of remote sensing images, and we make an exploratory study comparing a large number of collaboration strategies in order to better understand the interactions between these paradigms.

Original languageFrench
Pages (from-to)133-152
Number of pages20
JournalRevue d'Intelligence Artificielle
Volume31
Issue number1-2
DOIs
Publication statusPublished - 1 Jan 2017

    Fingerprint

Keywords

  • Classification
  • Remote sensing image analysis
  • Segmentation

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Cite this