February Fourier Talks 2014

Multiscale representations, in particular wavelets and over-complete representations, provide useful frameworks for processing and analyzing remote sensing imagery, e.g., for image registration and fusion. Such frameworks especially focus on computational speed and on the ability to handle multisource remote sensing data, such as multitemporal and multisensor images. In this talk, we will consider several multiscale representations and will investigate their applications when paired with various search strategies and similarity metrics, and when dealing with the registration and the fusion of large scenes acquired by several Earth Science space instruments, Landsat, MODIS and EO-1/ALI and Hyperion.