Search

The HYEP project (ANR HYEP 14-CE22-0016-01 Hyperspectral imagery for Environmental urban Planning - Mobility and Urban Systems Programme - 2014) confirmed the interest of a global approach to the urban environment by remote sensing and in particular by using hyperspectral imaging (HI). The interest of hyperspectral images lies in the range of information provided over wavelengths from 0.4 to 4 μm; they thus provide access to spectral quantities of interest and to chemical or biophysical parameters of the surface. HYEP's objective was to specify this and to propose a panel of methods and treatments taking into account the characteristics of other existing sensors in order to compare their performance. The developments carried out were applied and evaluated on hyperspectral airborne images acquired in Toulouse and Kaunas (Lithuania), also used to synthesize space systems: Sentinel-2, Hypxim/Biodiversity and Pleiades. Among the locks identified were those related to improving the spatial capabilities of the sensors and spatial scale changes, which were partially overcome by fusion and sharpening approaches, which proved to be successful. After a description of our hyperspectral data set acquired over Toulouse, an analysis is conducted on several existing and accessible spectral databases. Then, the chosen methods are presented. They include extraction, fusion and classification methods, which are then applied on our dataset synthetized at different spatial resolution to evaluate the benefits and the complementarity of hyperspectral imagery in comparison with other traditional sensors. Some specific applications are investigated of interest for urban planners: impervious soil map, vegetation species cartography and detection of solar panels. Finally, discussion and perspectives are presented. [ABSTRACT FROM AUTHOR]

A statistical model for detecting changes in remote sensing images has recently been proposed in (Prendes et al., 2014a,b). This model is sufficiently general to be used for homogeneous images acquired by the same kind of sensors (e.g., two optical images from Pléiades satellites, possibly with different acquisition conditions), and for heterogeneous images acquired by different sensors (e.g., an optical image acquired from a Pléiades satellite and a synthetic aperture radar (SAR) image acquired from a TerraSAR-X satellite). This model assumes that each pixel is distributed according to a mixture of distributions depending on the noise properties and on the sensor intensity responses to the actual scene. The parameters of the resulting statistical model can be estimated by using the classical expectation-maximization (EM) algorithm. The estimated parameters are finally used to learn the relationships between the images of interest, via a manifold learning strategy. These relationships are relevant for many image processing applications, particularly those requiring a similarity measure (e.g., image change detection and image registration). The main objective of this paper is to evaluate the performance of a change detection method based on this manifold learning strategy initially introduced in (Prendes et al., 2014a,b). This performance is evaluated by using results obtained with pairs of real optical images acquired from Pléiades satellites and pairs of optical and SAR images.

Remote sensing is a useful tool set for monitoring changes in forest ecosystems, particularly remote and otherwise inaccessible tracts of tropical forest. To revisit findings of earlier satellite-based studies of phenological variation in Amazonian forests, the current study focused on the variation of vegetation indices (VIs) of French Guiana. Specifically, 12 years of VI data which had been treated for bi-directional effects were extracted from the archives of 2 particular satellite sensors: the SPOT VEGETATION instrument, and MODIS (the MODerate resolution Imaging Spectroradiometer). At the scale of the entire territory of French Guiana, as well as at 4 sites across the territory, VI data displayed strong seasonal patterns, with the dry season months having significantly higher VI estimates than the wet season months. As VIs are considered proxies for photosynthetic activity, those seasonal variations would seem to indicate leaf-flushing across French Guiana's forests during the dry season months of September to October. [ABSTRACT FROM AUTHOR]

We propose a fully automated stereo pipeline for producing digital elevation models from Pléiades satellite images. The agility of the Pléiades satellites allows them to capture multiple views of the same target in a single pass, enabling new applications that exploit these quasi-simultaneous high-resolution images. Concretely the tri-stereo acquisition modality permits to reduce the occlusions and to cross-validate the observed points. This paper gives an overview of our pipeline, named s2p, and presents some digital elevation models and 3D point clouds built from Pléiades tri-stereo datasets. The data was provided by Airbus DS and the CNES through the RTU program. The particularity of the s2p algorithm is that it permits to use conventional stereo correlation tools, by performing a very precise image rectification of each stereo pair. Although the acquisition system does not fit the pinhole camera model, which is necessary to make the rectification possible, the errors due to the pinhole assumption were shown to be negligible for small enough image sizes. Thus, the whole image can be treated by cutting it into small tiles that are processed independently.

A statistical model for detecting changes in remote sensing images has recently been proposed in (Prendes et al., 2014, 2015). This model is sufficiently general to be used for homogeneous images acquired by the same kind of sensors (e.g., two optical images from Pléiades satellites, possibly with different acquisition conditions), and for heterogeneous images acquired by different sensors (e.g., an optical image acquired from a Pléiades satellite and a synthetic aperture radar (SAR) image acquired from a TerraSAR-X satellite). This model assumes that each pixel is distributed according to a mixture of distributions depending on the noise properties and on the sensor intensity responses to the actual scene. The parameters of the resulting statistical model can be estimated by using the classical expectation-maximization (EM) algorithm. The estimated parameters are finally used to learn the relationships between the images of interest, via a manifold learning strategy. These relationships are relevant for many image processing applications, particularly those requiring a similarity measure (e.g., image change detection and image registration). The main objective of this paper is to evaluate the performance of a change detection method based on this manifold learning strategy initially introduced in (Prendes et al., 2014, 2015). This performance is evaluated by using results obtained with pairs of real optical images acquired from Pléiades satellites and pairs of optical and SAR images. [ABSTRACT FROM AUTHOR]

We propose a fully automated stereo pipeline for producing digital elevation models from Pléiades satellite images. The agility of the Pléiades satellites allows them to capture multiple views of the same target in a single pass, enabling new applications that exploit these quasi-simultaneous high-resolution images. Concretely the tri-stereo acquisition modality permits to reduce the occlusions and to cross-validate the observed points. This paper gives an overview of our pipeline, named s2p, and presents some digital elevation models and 3D point clouds built from Pléiades tri-stereo datasets. The data was provided by Airbus DS and the CNES through the RTU program. The particularity of the s2p algorithm is that it permits to use conventional stereo correlation tools, by performing a very precise image rectification of each stereo pair. Although the acquisition system does not fit the pinhole camera model, which is necessary to make the rectification possible, the errors due to the pinhole assumption were shown to be negligible for small enough image sizes. Thus, the whole image can be treated by cutting it into small tiles that are processed independently. [ABSTRACT FROM AUTHOR]