Application to satellite images

The TAMI team has a research project in the field of satellite images with CNES since 2007. During the last years we have worked in different satellite image problems, most of them focused on the estimation of an elevation model from a couple of stereo images. During the previous project, conjointly with the CNES and the French research centre CMLA, we established a full stereo chain including the calibration, the matching process and the cloud visualization. The main task performed by our group was the design of a complete algorithm for looking for the correspondences between the two images of the couple and the design of validation criteria for theses correspondences. Some of the research results have been published in [1, 3] and  the patent [Pat3].

A new collaboration with the CNES started during this last project related to satellite Pleiades images. Our group started working in the fusion of panchromatic and chromatic data, acquired by the satellite at different resolutions, into a full color image. A first algorithm [2] was published using synthetic data and the application to real satellite PLEIADES is the objective of a publication in preparation.

figura_5

Figure 1. From a hight spatial resolution of a panchromatic image and a low spatial resolution of a multispectral image the objective is the reconstruction of the multispectral image at hight spatial resolution.

Main references:

[1] G. Blanchet, A. Buades, B. Coll,  J.M Morel, B. Rougé,  “Fattening free correlation”,   Journal of Mathematical Imaging and Vision, Vol. 41 (1), pp:109-121, 2011.

[2] J. Duran, A. Buades, B. Coll, and C. Sbert, A nonlocal variational model for pansharpening image fusion, SIAM J. Imaging Sci., vol. 7(2), pp. 761-796, 2014.

[3] A. Buades, G. Facciolo,  “Reliable local stereo matching for urban 3D reconstruction”, SIAM J. Imaging Sci., in press 2015.

[Pat3] G. Blanchet, A. Buades, B. Coll, J.M. Morel, B. Rougé, Procedimiento de establecimiento de correspondencia entre una primer imagen digital y una segunda imagen digital de una misma escena para la obtención de disparidades, Spanish patent Ref. P25155ES00,  Ref. PCT/ES2010/070813. Owner: UIB-CNES