Satellite Image Time Series Analysis Under Time Warping

[dhar]

Satellite Image Time Series Analysis Under Time Warping

[attachment=313]

Abstract

Satellite Image Time Series are becoming increasingly
available and will continue to do so in the coming years
thanks to the launch of space missions which aim at providing a
coverage of the Earth every few days with high spatial resolution.
In the case of optical imagery, it will be possible to produce land
use and cover change maps with detailed nomenclatures. However,
due to meteorological phenomena, such as clouds, these time series
will become irregular in terms of temporal sampling, and one will
need to compare time series with different lengths. In this paper,
we present an approach to image time series analysis which is
able to deal with irregularly sampled series and which also allows
the comparison of pairs of time series where each element of the
pair has a different number of samples.

INTRODUCTION

SATELLITE Image Time Series (SITS, for short) are a
precious resource for Earth monitoring. Current time series
have either high temporal resolution (Spot Vegetation, Modis)
or high spatial resolution (Landsat). In the coming years, both
high temporal and high spatial resolution SITS are going to
be widely available thanks to the European Space Agency s
(ESA) Sentinel program. Nowadays, satellites as the Taiwanese
Formosat-2 are already providing similar data, but with a limited
coverage of the Earth s surface and with only four spectral
bands.

DYNAMIC TIME WARPING FOR REMOTE SENSING

The previous subsection detailed the original definition of
DTW. However, its application to remote sensing requires to
make several choices and adaptations of the original method. In
this subsection, the following three points will be detailed:
A. the extension of DTW to multidimensional time series,
i.e., multispectral radiometric profiles of evolution;
B. the modification of DTW in order to avoid inconsistent
temporal distortions, e.g., forbidding the association of
winter sensed values with summer ones;
C. the handling cloud-contaminated images in the construction
of the sequences.

CONCLUSION

In this paper, we have introduced the DTW as a tool to deal
with two of the main issues raised by high temporal resolution
satellite image series, namely the irregular sampling in the
temporal dimension and the need for comparison of pairs of
time series having different number of samples.