Visual analytics for semantic queries of TerraSAR-X image content

Daniela Espinoza-Molina; Kevin Alonso; Mihai Datcu

doi:10.1117/12.2197588

Translator Disclaimer

15 October 2015 Visual analytics for semantic queries of TerraSAR-X image content

Daniela Espinoza-Molina, Kevin Alonso, Mihai Datcu

Author Affiliations +

Proceedings Volume 9642, SAR Image Analysis, Modeling, and Techniques XV; 96420E (2015) https://doi.org/10.1117/12.2197588
Event: SPIE Remote Sensing, 2015, Toulouse, France

Abstract

With the continuous image product acquisition of satellite missions, the size of the image archives is considerably increasing every day as well as the variety and complexity of their content, surpassing the end-user capacity to analyse and exploit them. Advances in the image retrieval field have contributed to the development of tools for interactive exploration and extraction of the images from huge archives using different parameters like metadata, key-words, and basic image descriptors. Even though we count on more powerful tools for automated image retrieval and data analysis, we still face the problem of understanding and analyzing the results. Thus, a systematic computational analysis of these results is required in order to provide to the end-user a summary of the archive content in comprehensible terms. In this context, visual analytics combines automated analysis with interactive visualizations analysis techniques for an effective understanding, reasoning and decision making on the basis of very large and complex datasets. Moreover, currently several researches are focused on associating the content of the images with semantic definitions for describing the data in a format to be easily understood by the end-user. In this paper, we present our approach for computing visual analytics and semantically querying the TerraSAR-X archive. Our approach is mainly composed of four steps: 1) the generation of a data model that explains the information contained in a TerraSAR-X product. The model is formed by primitive descriptors and metadata entries, 2) the storage of this model in a database system, 3) the semantic definition of the image content based on machine learning algorithms and relevance feedback, and 4) querying the image archive using semantic descriptors as query parameters and computing the statistical analysis of the query results. The experimental results shows that with the help of visual analytics and semantic definitions we are able to explain the image content using semantic terms and the relations between them answering questions such as what is the percentage of urban area in a region? or what is the distribution of water bodies in a city?

Citation Download Citation

Daniela Espinoza-Molina, Kevin Alonso, and Mihai Datcu "Visual analytics for semantic queries of TerraSAR-X image content", Proc. SPIE 9642, SAR Image Analysis, Modeling, and Techniques XV, 96420E (15 October 2015); https://doi.org/10.1117/12.2197588

ACCESS THE FULL ARTICLE