Share Email Print
cover

Proceedings Paper

Deep learning for multi-label land cover classification
Author(s): Konstantinos Karalas; Grigorios Tsagkatakis; Michalis Zervakis; Panagiotis Tsakalides
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Whereas single class classification has been a highly active topic in optical remote sensing, much less effort has been given to the multi-label classification framework, where pixels are associated with more than one labels, an approach closer to the reality than single-label classification. Given the complexity of this problem, identifying representative features extracted from raw images is of paramount importance. In this work, we investigate feature learning as a feature extraction process in order to identify the underlying explanatory patterns hidden in low-level satellite data for the purpose of multi-label classification. Sparse auto-encoders composed of a single hidden layer, as well as stacked in a greedy layer-wise fashion formulate the core concept of our approach. The results suggest that learning such sparse and abstract representations of the features can aid in both remote sensing and multi-label problems. The results presented in the paper correspond to a novel real dataset of annotated spectral imagery naturally leading to the multi-label formulation.

Paper Details

Date Published: 15 October 2015
PDF: 14 pages
Proc. SPIE 9643, Image and Signal Processing for Remote Sensing XXI, 96430Q (15 October 2015); doi: 10.1117/12.2195082
Show Author Affiliations
Konstantinos Karalas, Technical Univ. of Crete (Greece)
Foundation for Research and Technology, Heraklion (Greece)
Grigorios Tsagkatakis, Foundation for Research and Technology, Heraklion (Greece)
Michalis Zervakis, Technical Univ. of Crete (Greece)
Panagiotis Tsakalides, Technical Univ. of Crete (Greece)
Univ. of Crete (Greece)


Published in SPIE Proceedings Vol. 9643:
Image and Signal Processing for Remote Sensing XXI
Lorenzo Bruzzone, Editor(s)

© SPIE. Terms of Use
Back to Top