Interoperability of multiscale visual representations for satellite image big data

François Merciol (1) , Antoine Sauray (1) , Sébastien Lefèvre (1)

1 OBELIX - Environment observation with complex imagery

In this paper, we propose an interoperable solution for dealing with hierarchical representations of satellite images. Computationally intensive construction of the tree representation is performed on the server side, while the need for computational resources on the client side are greatly reduced (tree postprocessing or visualization). The proposed scheme is interoperable in the sense that it does not impose any constraints on the client environment, and API for C++, Java and Python languages are currently available. The communication is performed nodewise in a binary format using array structure for limiting the memory footprint.

Mots clés: Index Terms-Hierarchical image representation; Tree; Client- server framework; Interoperability

Domaines: Traitement des images [eess.IV]

François Merciol, Antoine Sauray, Sébastien Lefèvre. Interoperability of multiscale visual representations for satellite image big data. Conference on Big Data from Space (BiDS), 2016, Santa Cruz de Tenerife, Spain. ⟨hal-01253872⟩