Andreea Julea

Unsupervised Spatiotemporal Mining of Satellite Image Time Series Using Grouped Frequent Sequential Patterns

An important aspect of satellite image time series is the simultaneous access to spatial and temporal information. Various tools allow end users to interpret these data without having to browse the whole data set. In this paper, we intend to extract, in an unsupervised way, temporal evolutions at the pixel level and select those covering at least a minimum surface and having a high connectivity measure. To manage the huge amount of data and the large number of potential temporal evolutions, a new approach based on data-mining techniques is presented. We have developed a frequent sequential pattern extraction method adapted to that spatiotemporal context. A successful application to crop monitoring involving optical data is described. Another application to crustal deformation monitoring using synthetic aperture radar images gives an indication about the generic nature of the proposed approach.

Combining Airborne Photographs and Spaceborne SAR Data to Monitor Temperate Glaciers: Potentials and Limits

Monitoring temperate glacier activity has become more and more necessary for economical and security reasons and as an indicator of the local effects of global climate change. Remote sensing data provide useful information on such complex geophysical objects, but they require specific processing techniques to cope with the difficult context of moving and changing features in high-relief areas. This paper presents the first results of a project involving four laboratories developing and combining specific methods to extract information from optical and synthetic aperture radar (SAR) data. Two different information sources are processed, namely: 1) airborne photography and 2) spaceborne C-band SAR interferometry. The difficulties and limitations of their processing in the context of Alpine glaciers are discussed and illustrated on two glaciers located in the Mont-Blanc area. The results obtained by aerial triangulation techniques provide digital terrain models with an accuracy that is better than 30 cm, which is compatible with the computation of volume balance and useful for precise georeferencing and slope measurement updating. The results obtained by SAR differential interferometry using European Remote Sensing Satellite images show that it is possible to measure temperate glacier surface velocity fields from October to April in one-day interferograms with approximately 20-m ground sampling. This allows to derive ice surface strain rate fields required to model the glacier flow. These different measurements are complementary to results obtained during the summer from satellite optical data and ground measurements that are available only in few accessible points

Radar-Coding and Geocoding Lookup Tables for the Fusion of GIS and SAR Data in Mountain Areas

Synthetic aperture radar (SAR) image orthorectification induces an important alteration of information due to the side-looking geometry of SAR acquisition. In high-relief areas, the difficulty is increased by the foldover effect: The images acquired with low incidence angles cannot be registered by a bijective transformation like polynomial transformations, as usually proposed by conventional software. In this letter, a simple and efficient method, fitted to geocoded data and SAR images, is introduced to propose a generic coregistration tool that takes SAR geometry into account without requiring the exact sensor model, specific parameters, and precise navigation data. This method is based on a simulated SAR image and on the computation of lookup tables (LUTs) that represent the coordinate transform from one geometry to the other. Results are presented on a high-relief area in the Alps, where satellite and airborne SAR images are used for glacier evolution monitoring. A comparison to other sensor-independent approaches has been performed, showing that the proposed approach performs better in mountain areas. The resulting LUTs allow merging SAR data with the georeferenced data, either in ground geometry by orthorectifying the SAR information or in radar geometry by the inverse transformation, namely, radar-coding data from a geographic information system, to improve the analysis of SAR images and the result interpretation.

Combining optical and SAR data to monitor temperate glaciers

Monitoring temperate glaciers activity becomes more and more necessary for economical and security reasons and as an indicator of the local effects of global changes. This paper presents the beginning of a three year project which regroups four laboratories to develop and combine specific methods to extract information from optical and radar remote sensing data. Preliminary results are presented on three different information sources: airborne photography, space-borne multi-spectral images and SAR interferometry, which respectively allow the compution of high resolution DTM, the detection of glacial lakes and the measurement of glacier surface velocity. Results obtained on two glaciers located in the French Alps are compared and validated with ground measurments.

Mining pixel evolutions in satellite image time series for agricultural monitoring

In this paper, we present a technique to help the experts in agricultural monitoring, by mining Satellite Image Time Series over cultivated areas. We use frequent sequential patterns extended to this spatiotemporal context in order to extract sets of connected pixels sharing a similar temporal evolution. We show that a pixel connectivity constraint can be partially pushed to prune the search space, in conjunction with a support threshold. Together with a simple maximality constraint, the method reveals meaningful patterns in real data.

Extraction of frequent grouped sequential patterns from Satellite Image Time Series

This paper presents an original data mining approach for extracting pixel evolutions and sub-evolutions from Satellite Image Time Series. These patterns, called frequent grouped sequential patterns, represent the (sub-)evolutions of pixels over time, and have to satisfy two constraints: firstly to correspond to at least a given minimum surface and secondly to be shared by pixels that are sufficiently connected. These spatial constraints are actively used to face large data volumes and to select evolutions making sense for end-users. Successful experiments on an optical and a radar SITS are presented.

Polsar RADARSAT-2 Satellite Image Time Series mining over the Chamonix Mont-Blanc test site

This paper presents a data mining approach for describing Satellite Image Time Series (SITS) spatially and temporally. It relies on pixel-based evolution and sub-evolution extraction. These evolutions, namely the {frequent grouped sequential patterns}, are required to cover a minimum surface and to affect pixels that are sufficiently connected. These spatial constraints are actively used to face large data volumes and to select evolutions making sense for end-users. In this paper, a specific application to fully polarimetric SAR image time series is presented. Experiments performed on a RADARSAT-2 SITS covering the Chamonix Mont-Blanc test-site are used to illustrate the proposed approach.

High Resolution SAR Interferometry: Influence of Local Topography in the Context of Glacier Monitoring

SAR interferometric data offers the opportunity to measure temperate glacier surface topography and displacement between the two acquisitions. Recently, reliable estimates of the phase gradient given by interferogram local frequencies become mandatory with the increase of the SAR resolution. In this paper, an original 2-step method for estimating local frequencies is proposed. The 2D phase signal is considered to have two deterministic components corresponding to low-resolution fringes and high-resolution patterns due for instance to the micro-relief. The first step of the proposed algorithm consists in the low-resolution phase flattening. In the second step the local high-resolution frequencies are estimated from the phase auto-correlation functions computed on adaptive neighborhoods using only the pixels which belong to the same HR spatial feature and respect the ”local stationarity” hypothesis. Results with both real ERS 1/2 tandem and simulated TerraSAR-X interferograms are presented to illustrate the potential of the proposed method.

Connectivity constraint-based sequential pattern extraction from Satellite Image Time Series (SITS)

The temporal evolution of pixel values in Satellite Image Time Series (SITS) is considered as criterion for the characterization, discrimination and identification of terrestrial objects and phenomena. Due to the exponential behavior of sequences number with specialization, Sequential Data Mining (SDM) techniques need to be applied. The huge search and solution spaces imply the use of constraints according to the user’s knowledge, interest and expectation. The spatial aspect of the data was taken into account by the introduction of connectivity measures that characterize the pixels tendency to form objects. These measures can highlight stratifications in data structure, can be useful for shape recognition and offer a base for post-processing operations similar to those from mathematical morphology (dilation, erosion etc.). The conjunction of corresponding Connectivity Constraints (CC) with the Support Constraint (SC) leads to the extraction of Grouped Frequent Sequential Patterns (GFSP), a concept with proved capability for preliminary description and localization of terrestrial events. This work is focused on efficient SITS extraction of evolutions that fulfill SC and CC. Different types of extractions using anti-monotone constraints are analyzed. Experiments performed on two interferometric SITS are used to illustrate the potential of the approach to find interesting evolution patterns.

Urban area monitoring by sequential patterns extracted from multi-temporal satellite data using connectivity measures

The temporal evolution of pixel values in Satellite Image Time Series (SITS) is considered criterion for the characterization, discrimination and identification of terrestrial objects and phenomena. Due to the exponential behavior of sequences number with specialization, Sequential Data Mining techniques need to be applied. The spatial aspect of the data was taken into account by the introduction of connectivity measures that characterize the pixels tendency to form objects. The conjunction of corresponding Connectivity Constraints (CC) with the Support Constraint (SC) leads to the extraction of Grouped Frequent Sequential Patterns (GFSP), a concept with proved capability for preliminary description and localization of terrestrial events. This work is focused on efficient SITS extraction of evolutions that fulfil SC and CC. Experiments performed on Bucharest urban interferometric SITS are used to illustrate the potential of the approach to find interesting evolution patterns.