Assessment of Deep Learning Techniques for Land Use Land Cover Classification in Southern New Caledonia

Type Article
Date 2021-06
Language English
Author(s) Rousset Guillaume1, 2, Despinoy MarcORCID1, Schindler KonradORCID3, Mangeas MorganORCID4
Affiliation(s) 1 : ESPACE-DEV, Univ New Caledonia, Univ Montpellier, IRD, Univ Antilles, Univ Guyane, Univ Réunion, 98800 New Caledonia, France
2 : ISEA, Univ New Caledonia, 98800 New Caledonia, France
3 : Photogrammetry and Remote Sensing, ETH Zurich, 8093 Zurich, Switzerland
4 : Institut de Recherche pour le Développement, UMR 9220 ENTROPIE (Institut de Recherche pour le Développement, Université de la Réunion, IFREMER, Université de la Nouvelle-Calédonie, Centre National de la Recherche Scientifique), 98800 New Caledonia, France
Source Remote Sensing (2072-4292) (MDPI AG), 2021-06 , Vol. 13 , N. 12 , P. 2257 (22p.)
DOI 10.3390/rs13122257
Keyword(s) New Caledonia, remote sensing, land use, land cover, deep learning, XGBoost, neural network, neo-channels
Abstract

Land use (LU) and land cover (LC) are two complementary pieces of cartographic information used for urban planning and environmental monitoring. In the context of New Caledonia, a biodiversity hotspot, the availability of up-to-date LULC maps is essential to monitor the impact of extreme events such as cyclones and human activities on the environment. With the democratization of satellite data and the development of high-performance deep learning techniques, it is possible to create these data automatically. This work aims at determining the best current deep learning configuration (pixel-wise vs. semantic labelling architectures, data augmentation, image prepossessing, …), to perform LULC mapping in a complex, subtropical environment. For this purpose, a specific data set based on SPOT6 satellite data was created and made available for the scientific community as an LULC benchmark in a tropical, complex environment using five representative areas of New Caledonia labelled by a human operator: four used as training sets, and the fifth as a test set. Several architectures were trained and the resulting classification was compared with a state-of-the-art machine learning technique: XGboost. We also assessed the relevance of popular neo-channels derived from the raw observations in the context of deep learning. The deep learning approach showed comparable results to XGboost for LC detection and over-performed it on the LU detection task (61.45% vs. 51.56% of overall accuracy). Finally, adding LC classification output of the dedicated deep learning architecture to the raw channels input significantly improved the overall accuracy of the deep learning LU classification task (63.61% of overall accuracy). All the data used in this study are available on line for the remote sensing community and for assessing other LULC detection techniques.

Full Text
File Pages Size Access
Publisher's official version 22 7 MB Open access
Supplementary File 1 33 MB Open access
Top of the page