Abstract
Land-cover map production using remote-sensing imagery is governed by data availability. In our case, data sources are two-fold: on one hand, optical data provided regularly by satellites such as Sentinel-2, and on the other hand, reference data which allow calibrating mapping methods or validating the results. The lengthy delays due to reference data collection and cleansing are one of the main issues for applications. In this work, the use of Optimal Transport (OT) is proposed. OT is a Domain Adaptation method that uses past data, both images and reference data, to produce the land-cover map of the current period without updated reference data. Seven years of Formosat-2 image time series and the corresponding reference data are used to evaluate two OT algorithms: conventional EMD transport and regularized transport based on the Sinkhorn distance. The contribution of OT to a classification fusion strategy is also evaluated. The results show that with a 17-class nomenclature the problem is too complex for the Sinkhorn algorithm, which provides maps with an Overall Accuracy (OA) of 30%. In contrast, with the EMD algorithm, an OA close to 70% is obtained. One limitation of OT is the number of classes that can be considered at the same time. Simplification schemes are proposed to reduce the number of classes to be transported. Cases of improvement are shown when the problem is simplified, with an improvement in OA varying from 5% and 20%, producing maps with an OA near 79%. As several years are available, the OT approaches are compared to standard fusion schemes, like majority voting. The gain in voting strategies with OT use is lower than the gain obtained with standard majority voting (around 5%).
Highlights
A land-cover map is an image where each pixel value corresponds to a land-cover class label
The contribution of the majority voting using the maps produced with the Earth Mover’s Distance (EMD)-PerClass approach is compared to majority voting using the naive maps
Considering 2007 as source domain, the different pairs of years are [(2007, 2008), (2007, 2009), · · ·, (2007, 2013)], the Overall Accuracy (OA) obtained on the 10 draws for each pair of years are averaged and the 90% OA confidence interval is computed
Summary
A land-cover map is an image where each pixel value corresponds to a land-cover class label. Recent decades have seen the launch of numerous satellites dedicated to Earth monitoring which provide high-quality data for land-cover mapping, such as the SPOT and Landsat families and more recently the Sentinel-2 constellation. Classification algorithms based on machine learning methodologies are often applied, which consist of a set of decision rules assigning a label to each pixel. Maximum Likelihood classifiers have been used for land-cover mapping [10], but they do not behave well in high dimensional feature spaces as those encountered with multispectral image time series. SVM and RF can handle data sets with large numbers of features, such as the various spectral bands for the different dates of satellite image time series, and with many training samples covering a wide geographical extent. In addition to image time series, high-quality labelled reference data are an essential element for an efficient classification [11]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
