This work investigates the use of deep fully convolutional neural networks (DFCNN) for pixel-wise scene labeling of Earth Observation images. Especially, we train a variant of the SegNet architecture on remote sensing data over an urban area and study different strategies for performing accurate semantic segmentation. Our contributions are the following: 1) we transfer efficiently a DFCNN from generic everyday images to remote sensing images; 2) we introduce a multi-kernel convolutional layer for fast aggregation of predictions at multiple scales; 3) we perform data fusion from heterogeneous sensors (optical and laser) using residual correction. Our framework improves state-of-the-art accuracy on the ISPRS Vaihingen 2D Semantic Labeling dataset.
@article{arxiv.1609.06846,
title = {Semantic Segmentation of Earth Observation Data Using Multimodal and Multi-scale Deep Networks},
author = {Nicolas Audebert and Bertrand Le Saux and Sébastien Lefèvre},
journal= {arXiv preprint arXiv:1609.06846},
year = {2016}
}
Comments
Asian Conference on Computer Vision (ACCV16), Nov 2016, Taipei, Taiwan