Gabriele Cavallaro

Research Topic: Analysis of Very High Resolution (VHR) remote sensing images. Extraction of spatial information of the surveyed scene by implementing attribute profiles computed on a hierarchical data representation. Self Dual Attribute Profiles are exploited and developed for classification purposes.

Parallel project: High Performance Computing of Hyperspectral remote sensing images. GPU implementation of unmixing algorithms.

Keywords: remote sensing, image processing, pattern recognition, unmixing, hyperspectral/multispectral remote sensing images, high performance computing, gpgpu.

Data sets: hyperspectral images: AVIRIS, ROSIS, Hyperion, HYDICE. Multispectral and Panchromatic images: QuickBird, IKONOS.

Software:

  • Visual Studio (C, C++, OpenCL)
  • MATLAB
  • IDL/ENVI