The general problem of matching structures is very pervasive in computer vision and image process... more The general problem of matching structures is very pervasive in computer vision and image processing. The research presented here tackles the problem of object matching in a very general perspective. It is formulated for the matching of surfaces. It applies to objects having small or large deformation and arbitrary topological changes. The process described hinges on a geodesic distance equation for a family of curves or surfaces embedded in the graph of a cost function. This geometrical approach to object matching has the advantage that the similarity criterion can be used to define the shape of the cost function. Matching paths are computed on the cost manifolds using distance maps. These distance maps are generated by solving a general partial differential equation which is a generalization of the geodesic dis-tance evolution scheme introduced by R. Kimmel, A. Amir, and A. F. Bruckstein (1995, IEEE Trans. Pattern Anal. Mach. Intell. 17, 635–640). An Eulerian level-set formulation...
The NOAA-AVHRR sensor provides daily acquisitions which may be used for vegetation growth monitor... more The NOAA-AVHRR sensor provides daily acquisitions which may be used for vegetation growth monitoring. However, the coarse resolution of these data causes a problem of pixel heterogeneity. The pixel radiometric value is, in fact, a composition of individual responses from the different land covers found within the pixel’s surface. In this paper we assume a linear relation. The simultaneous use of NOAA-AVHRR temporal series and of high spatial resolu- tion ground data makes it possible to estimate the pure NOAA reflectances in the visible and near infrared wavelengths for each vegetation type. These val- ues are then combined to compute the NDVI temporal profile which describes the seasonal cycle of the studied crops. The proposed method is applied on the region of Chartres in order to describe the major crops temporal behavior.
Cet article propose une methode de prevision a court terme des precipitations, basee sur deux typ... more Cet article propose une methode de prevision a court terme des precipitations, basee sur deux types d'images radar avec des resolu-tions spatiales complementaires. Cette methode est implementee grâce a l'elaboration d'une version multi-echelles d'un algorithme operationnel. Differentes variantes de la methode permettent egalement d'augmenter l'horizon de prevision. Une evaluation quantitative permet de valider l'ap-proche est de demontrer sont potentiel. Abstract-The paper describes a rain nowcasting method, based on multi-resolution radar images. A multi-scale version of an operational algorithm is given. Alternatives to this method are also proposed to increase the temporal horizon of the forecasts. A quantitative evaluation validates the approach.
Satellite data are daily acquired over Black Sea and allow visualizing the surface circulation an... more Satellite data are daily acquired over Black Sea and allow visualizing the surface circulation and its meso-scale structures at the spatial resolution and time frequency of the sensor. The remotely-sensed images are in fact the only continuous source of information at fine scales greater for analysing eddies, jets, filaments, mushroom-shaped vortices... A number of applications, such as oil spill monitoring for instance, also require being able to forecast the surface circulation at short temporal horizon. This nowcasting issue includes two major components: a first module is in charge of continuously estimating the meso-scale structures of the surface dynamics and a second module is in charge of forecasting the future image data at a given temporal horizon. The presentation summarizes these two modules. It also provides results obtained with NOAA-AVHRR data acquired over Black Sea. The estimation module relies on a sliding-window approach, which processes the last few images and al...
Proceedings. 2005 IEEE International Geoscience and Remote Sensing Symposium, 2005. IGARSS '05., 2005
This paper aims at monitoring agricultural areas using time series of satellite images, in order ... more This paper aims at monitoring agricultural areas using time series of satellite images, in order to obtain as early as possible in the year a classification of cultivated areas, as well as a characterisation of the different phenological behaviors for cultivated and natural land covers. For that purpose a learning phase has been performed, to provide a mathematical characterization of
Proceedings of 13th International Conference on Pattern Recognition, 1996
Motion study in computer vision highly depends on the type of imagery and on the kind of the stud... more Motion study in computer vision highly depends on the type of imagery and on the kind of the studied objects. A popular approach of non-rigid motion supposes that the objects are continuously and locally deformed. Local features, such as curvature extrema, are then often used. However, these features can not always be accurately computed, and motion may involve large deformation of the object between two consecutive temporal occurrences. In these cases, a real need exists for an approach that does not rely on local features. That study of motion requires additional information. We introduce a geometrical evolution model that enables one to generate a surface interpolating successive contours of the object during its temporal evolution. This geometrical model may be viewed as a simplification of a true physical model of motion. This approach is particularly well suited to remote sensed data: structures of interest do not have a well defined shape, and the temporal resolution may be poor, involving large deformation. The model is successfully applied to vortex tracking on sea color and meteorologic images.
The general problem of matching structures is very pervasive in computer vision and image process... more The general problem of matching structures is very pervasive in computer vision and image processing. The research presented here tackles the problem of object matching in a very general perspective. It is formulated for the matching of surfaces. It applies to objects having small or large deformation and arbitrary topological changes. The process described hinges on a geodesic distance equation for a family of curves or surfaces embedded in the graph of a cost function. This geometrical approach to object matching has the advantage that the similarity criterion can be used to define the shape of the cost function. Matching paths are computed on the cost manifolds using distance maps. These distance maps are generated by solving a general partial differential equation which is a generalization of the geodesic dis-tance evolution scheme introduced by R. Kimmel, A. Amir, and A. F. Bruckstein (1995, IEEE Trans. Pattern Anal. Mach. Intell. 17, 635–640). An Eulerian level-set formulation...
The NOAA-AVHRR sensor provides daily acquisitions which may be used for vegetation growth monitor... more The NOAA-AVHRR sensor provides daily acquisitions which may be used for vegetation growth monitoring. However, the coarse resolution of these data causes a problem of pixel heterogeneity. The pixel radiometric value is, in fact, a composition of individual responses from the different land covers found within the pixel’s surface. In this paper we assume a linear relation. The simultaneous use of NOAA-AVHRR temporal series and of high spatial resolu- tion ground data makes it possible to estimate the pure NOAA reflectances in the visible and near infrared wavelengths for each vegetation type. These val- ues are then combined to compute the NDVI temporal profile which describes the seasonal cycle of the studied crops. The proposed method is applied on the region of Chartres in order to describe the major crops temporal behavior.
Cet article propose une methode de prevision a court terme des precipitations, basee sur deux typ... more Cet article propose une methode de prevision a court terme des precipitations, basee sur deux types d'images radar avec des resolu-tions spatiales complementaires. Cette methode est implementee grâce a l'elaboration d'une version multi-echelles d'un algorithme operationnel. Differentes variantes de la methode permettent egalement d'augmenter l'horizon de prevision. Une evaluation quantitative permet de valider l'ap-proche est de demontrer sont potentiel. Abstract-The paper describes a rain nowcasting method, based on multi-resolution radar images. A multi-scale version of an operational algorithm is given. Alternatives to this method are also proposed to increase the temporal horizon of the forecasts. A quantitative evaluation validates the approach.
Satellite data are daily acquired over Black Sea and allow visualizing the surface circulation an... more Satellite data are daily acquired over Black Sea and allow visualizing the surface circulation and its meso-scale structures at the spatial resolution and time frequency of the sensor. The remotely-sensed images are in fact the only continuous source of information at fine scales greater for analysing eddies, jets, filaments, mushroom-shaped vortices... A number of applications, such as oil spill monitoring for instance, also require being able to forecast the surface circulation at short temporal horizon. This nowcasting issue includes two major components: a first module is in charge of continuously estimating the meso-scale structures of the surface dynamics and a second module is in charge of forecasting the future image data at a given temporal horizon. The presentation summarizes these two modules. It also provides results obtained with NOAA-AVHRR data acquired over Black Sea. The estimation module relies on a sliding-window approach, which processes the last few images and al...
Proceedings. 2005 IEEE International Geoscience and Remote Sensing Symposium, 2005. IGARSS '05., 2005
This paper aims at monitoring agricultural areas using time series of satellite images, in order ... more This paper aims at monitoring agricultural areas using time series of satellite images, in order to obtain as early as possible in the year a classification of cultivated areas, as well as a characterisation of the different phenological behaviors for cultivated and natural land covers. For that purpose a learning phase has been performed, to provide a mathematical characterization of
Proceedings of 13th International Conference on Pattern Recognition, 1996
Motion study in computer vision highly depends on the type of imagery and on the kind of the stud... more Motion study in computer vision highly depends on the type of imagery and on the kind of the studied objects. A popular approach of non-rigid motion supposes that the objects are continuously and locally deformed. Local features, such as curvature extrema, are then often used. However, these features can not always be accurately computed, and motion may involve large deformation of the object between two consecutive temporal occurrences. In these cases, a real need exists for an approach that does not rely on local features. That study of motion requires additional information. We introduce a geometrical evolution model that enables one to generate a surface interpolating successive contours of the object during its temporal evolution. This geometrical model may be viewed as a simplification of a true physical model of motion. This approach is particularly well suited to remote sensed data: structures of interest do not have a well defined shape, and the temporal resolution may be poor, involving large deformation. The model is successfully applied to vortex tracking on sea color and meteorologic images.
Uploads
Papers by Isabelle HERLIN