Abstract
Classification of very high resolution (VHR) satellite images faces two major challenges: 1) inherent low intra-class and high inter-class spectral similarities and 2) mismatching resolution of available bands. Conventional methods have addressed these challenges by adopting separate stages of image fusion and spatial feature extraction steps. These steps, however, are not jointly optimizing the classification task at hand. We propose a single-stage framework embedding these processing stages in a multiresolution convolutional network. The network, called FuseNet, aims to match the resolution of the panchromatic and multispectral bands in a VHR image using convolutional layers with corresponding downsampling and upsampling operations. We compared FuseNet against the use of separate processing steps for image fusion, such as pansharpening and resampling through interpolation. We also analyzed the sensitivity of the classification performance of FuseNet to a selected number of its hyperparameters. Results show that FuseNet surpasses conventional methods.
Original language | English |
---|---|
Title of host publication | 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings |
Publisher | IEEE |
Pages | 2091-2094 |
Number of pages | 4 |
Volume | 2018-July |
ISBN (Electronic) | 9781538671504 |
DOIs | |
Publication status | Published - 31 Oct 2018 |
Event | 38th IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018: Observing, Understanding and Forcasting the Dynamics of Our Planet - Feria Valencia Convention & Exhibition Center, Valencia, Spain Duration: 22 Jul 2018 → 27 Jul 2018 Conference number: 38 https://www.igarss2018.org/ |
Conference
Conference | 38th IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 |
---|---|
Abbreviated title | 2018 |
Country/Territory | Spain |
City | Valencia |
Period | 22/07/18 → 27/07/18 |
Internet address |
Keywords
- Convolutional networks
- Deep learning
- Image fusion
- Land cover classification
- VHR image