Deep learning for semantic segmentation of remote sensing images with rich spectral content.
Amina Ben HamidaAlexandre BenoîtPatrick LambertLouis KleinChokri Ben AmarNicolas AudebertSébastien LefèvrePublished in: IGARSS (2017)
Keyphrases
- remote sensing images
- deep learning
- weakly supervised
- remote sensing
- multispectral
- hyperspectral
- superpixels
- unsupervised learning
- machine learning
- image data
- semi supervised
- high quality
- data sets
- natural images
- conditional random fields
- image registration
- object segmentation
- long range
- object class
- similarity measure
- image analysis