DEEP LEARNING FOR SEMANTIC SEGMENTATION OF REMOTE SENSING IMAGES WITH RICH SPECTRAL CONTENT

Amina Ben Hamida , Alexandre Benoit , P. Lambert (1) , L Klein , Chokri Ben Amar , Nicolas Audebert (2) , Sébastien Lefèvre (3)

1 LISTIC - Laboratoire d'Informatique, Systèmes, Traitement de l'Information et de la Connaissance
2 ONERA - The French Aerospace Lab [Palaiseau]
3 VALORIA - Laboratoire de Recherche en Informatique et ses Applications de Vannes et Lorient

With the rapid development of Remote Sensing acquisition techniques, there is a need to scale and improve processing tools to cope with the observed increase of both data volume and richness. Among popular techniques in remote sensing, Deep Learning gains increasing interest but depends on the quality of the training data. Therefore, this paper presents recent Deep Learning approaches for fine or coarse land cover semantic segmentation estimation. Various 2D architectures are tested and a new 3D model is introduced in order to jointly process the spatial and spectral dimensions of the data. Such a set of networks enables the comparison of the different spectral fusion schemes. Besides, we also assess the use of a " noisy ground truth " (i.e. outdated and low spatial resolution labels) for training and testing the networks.

Mots clés: Multispectral; Deep Learning; Semantic Segmentation; Noisy Training; Index Terms— Remote Sensing

Domaines: Traitement du signal et de l'image [eess.SP]; Intelligence artificielle [cs.AI]; Vision par ordinateur et reconnaissance de formes [cs.CV]; Apprentissage [cs.LG]; Recherche d'information [cs.IR]; Réseau de neurones [cs.NE]; Sciences de l'environnement

Amina Ben Hamida, Alexandre Benoit, P. Lambert, L Klein, Chokri Ben Amar, et al.. DEEP LEARNING FOR SEMANTIC SEGMENTATION OF REMOTE SENSING IMAGES WITH RICH SPECTRAL CONTENT. IEEE International Geoscience and Remote Sensing Symposium, Jul 2017, Fort Worth, United States. ⟨hal-01654187⟩