Papers by Eliane Cubero-Castan
Franck Sery
: The multilook polarimetric maximum likelihood classifier based on the Wishart distribution supp... more : The multilook polarimetric maximum likelihood classifier based on the Wishart distribution supposes no variation of the backscattering of the underlying scene. For clutters verifying the product model », we here present the use of a K-distribution and compare this classifier to the one based on the Wishart distribution. A simple way to obtain a full polarimetric filter by filtering a set of adequate powers is also given. We show how filtering and segmentation of the raw data improve the classification results. KEY WORDS : SAR images, multitemporal and multisource classification, segmentation, polarimetry, texture, speckle filtering 1. INTRODUCTION Statistical distributions of fully developped speckle in regions of stable underlying reflectivity have been established theoretically for full polarimetric data [1]. These distributions have been used to define maximum likelihood classifiers [2,3]. The optimality of a maximum likelihood classifier is related to the goodness of fit of the r...
Operational scheduling of Direct Tasking innovative concept to improve reactivity on earth observation system
14th International Conference on Space Operations, 2016
Edge Detection in Radar Images using Recursive Filters
Edge detection is a fundamental issue in image analysis. A common approach is to identify edges a... more Edge detection is a fundamental issue in image analysis. A common approach is to identify edges as local maxima of the gradient magnitude in the gradient direction. In radar images, the presence of a non-Gaussian multiplicative noise known as speckle makes standard methods ...
Contrast Invariant Image Intersection
... 7 Page 8. References 1] PE Anuta, Spatial Registration of Multispectral and Multitemporal Dig... more ... 7 Page 8. References 1] PE Anuta, Spatial Registration of Multispectral and Multitemporal Digital Imagery Using Fast Fourier Transform Techniques. ... 7] J. Hsieh, H. Liao, K. Fan, M. Ko. A Fast Algorithm for Image Registration without Predeter-mining Correspondences. ...
Updating cartographic models by Spot images interpretation
This paper presents the first results of a study devoted to the update of cartographic models wit... more This paper presents the first results of a study devoted to the update of cartographic models with multispectral Spot images. With the current proliferation of GIS, this updating problem will be of prime importance in the next years. We first present the general issue and the tested application, which is limited to surfacic entities, small number of model classes, and
<title>Neural approach for satellite image registration and pairing segmented areas</title>
Image and Signal Processing for Remote Sensing II, 1995
Evaluation on SPOT data of classification algorithms based on Markovian modelization
1995 International Geoscience and Remote Sensing Symposium, IGARSS '95. Quantitative Remote Sensing for Science and Applications, 1995
CNES (French Space Agency) has developed a research program related to SPOT imagery to deal with ... more CNES (French Space Agency) has developed a research program related to SPOT imagery to deal with cartography topics. Many studies, conducted with different laboratories, are intended to work on remote sensing data. The main purpose of the present research is information extraction (network extraction, urban area extraction, segmentation, etc.) One of these studies, made in collaboration with INRIA Sophia Antipolis,
International charter 'space and major disasters' status report
Proceedings. 2005 IEEE International Geoscience and Remote Sensing Symposium, 2005. IGARSS '05., 2005
Disasters' celebrates in October 2005 its 5th anniversary of operations. The Charter has bee... more Disasters' celebrates in October 2005 its 5th anniversary of operations. The Charter has been activated for seventy-two times to respond to mostly natural but also technological disasters that struck communities all over the world. The Charter membership has grown ...
<title>Applying co-operative operators for urban-area detection using SPOT imagery</title>
Image and Signal Processing for Remote Sensing III, 1996
Comments on (quote)Geodesic saliency of watershed contours and hierarchical segmentation(quote) [and reply]
IEEE Transactions on Pattern Analysis and Machine Intelligence, 1998
<title>Robust 2D phase unwrapping based on multiresolution</title>
Microwave Sensing and Synthetic Aperture Radar, 1996
ABSTRACT To model the impact of the landcover on the climatic and hydrological cycles, an exact k... more ABSTRACT To model the impact of the landcover on the climatic and hydrological cycles, an exact knowledge of two facts is necessary: the landuse and the amount of biomass. The study area for this work is the catchment of the Ammer river, which covers about 1200 km2 in the Bavarian Alpine Foreland. Optical remote sensing data are proved to provide good information sources to derive landuse classifications for large areas. But due to the fact that commonly used classification algorithms are solely based on the spectral information, this often leads to misclassifications, because different classes can show similar spectral signatures. To derive a sufficient landuse classification for the testsite purely from remote sensing data, shows up to be difficult. Due to the increasing cloudiness at the border of the alps, the use of multitemporal data is limited. Moreover, the diverse structure of the testsite limits the use of Bayes- theory based classification approach, non-spectral geographical ancillary data, such as climatic and soil data are integrated. Rules for influencing parameters were derived and taken into account in the classification procedure. The developed approach is based on the possibility theory and fuzzy subsets. The results are verified with a digital ground truth map and show a substantially increase of the classification accuracies. To calculate the evaporation, besides the landuse pattern the development of the vegetation cover is of importance. To monitor the vegetation dynamics, multitemporal optical data are not available. Therefore, ERS-SAR radar data are used for this task. Since grassland is the dominating agricultural landuse, investigations were made on the utilization of radar data for the determination of the temporal development of grassland biomass. It is shown that there is a correlation between the signal intensity and the vegetation height of meadows. Due to the fact that the height highly correlates with the biomass, the grassland biomass can be estimated in the testsite.© (1997) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.
Different approaches to multiedge detection in SAR images
IGARSS'97. 1997 IEEE International Geoscience and Remote Sensing Symposium Proceedings. Remote Sensing - A Scientific Vision for Sustainable Development, 1997
E-mail: Eliane. Cubero-Castan@cnes.fr Fax: (33) 5-61-27-3 1-67 ... Abstract -- Edge detection is ... more E-mail: Eliane. Cubero-Castan@cnes.fr Fax: (33) 5-61-27-3 1-67 ... Abstract -- Edge detection is a fundamental issue in image analysis. The presence of speckle, which can be modelled as a strong multiplicative noise, complicates edge detection in Synthetic Aperture Radar ...
<title>Urban area extraction from a satellite image</title>
Image and Signal Processing for Remote Sensing II, 1995
ABSTRACT
1997 IEEE International Conference on Acoustics, Speech, and Signal Processing, 1997
Edge detection is a fundamental issue in image analysis. Due to the presence of speckle, which ca... more Edge detection is a fundamental issue in image analysis. Due to the presence of speckle, which can be modelled as a strong multiplicative noise, edge detection in Synthetic Aperture Radar (SAR) images is very di cult and methods developed for optical images are ine cient. We here propose a new edge detector for SAR images which is optimum in the MSSE sense for a stochastic multiedge model. It computes a normalized Ratio Of Exponentially Weighted Averages (ROEWA) on opposite sides of the central pixel. This is done in the horizontal and vertical direction, and the magnitude of the two components yields an edge strength map. Thresholding of the edge strength map and postprocessing to eliminate false edges are brie y discussed. We present results on simulated SAR images and ERS1 data. c IEEE 1997
Disaster Response in Africa by the International Charter
The International Charter 'Space and Major Disasters' is an initiative on the part of its... more The International Charter 'Space and Major Disasters' is an initiative on the part of its member European, French, Canadian, Indian, U.S. and Argentinean space agencies to pool together their satellite resources to provide data in the event of a natural or technological disaster. In the first four years of its operation, the Charter was in fact activated six times to
<title>Updating cartographic models by Spot images interpretation</title>
Image and Signal Processing for Remote Sensing, 1994
IGARSS '96. 1996 International Geoscience and Remote Sensing Symposium, 1996
Radar images are inherently degraded by a strong, multiplicative noise known as speckle. The most... more Radar images are inherently degraded by a strong, multiplicative noise known as speckle. The most frequently used speckle filters are adaptive in the sense that the filtering operation depends on estimations of local statistics calculated on a neighbourhood of the considered pixel. The choice of the neighbourhood is consequently an important issue. In this paper we introduce a new method which uses segmentations obtained prior to filtering. This region-based approach is compared to versions using sliding windows. The study is limited to agricultural scenes composed of distinct parcels of relatively homogeneous reflectivity. Without loss of generality, we have restricted ourselves to the LMMSE filter of Kuan et al.
Proceedings of 3rd IEEE International Conference on Image Processing, 1996
Segmentation plays a key role in many image processing applications, including SAR image analysis... more Segmentation plays a key role in many image processing applications, including SAR image analysis. Given a highquality segmentation, important tasks such as adaptive noise filtering and classification can be greatly simplified. However, the presence of a strong, multiplicative noise known as speckle makes edge detection in SAR images very difficult. A new, hybrid segmentation technique has permitted us to obtain segmentations of relatively good quality prior to filtering. This has led us to develop a region-based version of the LMMSE speckle filter. The segmentations may further be used for the classification of raw or filtered images. In this paper we give an overview of the entire scheme, which segments, filters and classifies SAR images. The study is limited to agricultural scenes composed of distinct parcels of relatively homogeneous reflectivity.
<title>Feature extraction and pattern classification for remotely sensed data analysis by a modular neural system</title>
Image and Signal Processing for Remote Sensing, 1994
<title>Multisource classification of SAR images with the use of segmentation, polarimetry, texture, and multitemporal data</title>
Image and Signal Processing for Remote Sensing III, 1996
Uploads
Papers by Eliane Cubero-Castan