Creating and sharing knowledge for telecommunications

Project: High-performance Onboard Hyperspectral Unmixing Systems

Acronym: HoHus
Main Objective:
Hyperspectral imaging has recently emerged as a powerful remote sensing technology.
The huge data volumes acquired by these sensors put stringent requirements on data processing techniques as most of the remote sensing applications require an immediate response.
The project aims at the development and implementation in GPU and FPGA hardware of the main steps of the unmixing techniques in a parallel fashion.
The full unmixing system supported by specialized hardware may be suitable in the future for real-time onboard processing.
Reference: P01133
Funding: IT/LA
Start Date: 01-10-2011
End Date: 01-10-2013
Team: José Manuel Peixoto do Nascimento, José Manuel Bioucas Dias, Vitor Manuel Mendes da Silva, Jose Manuel Rodriguez Alves
Groups: Pattern and Image Analysis – Lx, Multimedia Signal Processing – Co
Partners: António Plaza - University of Extremadura, Spain
Local Coordinator: José Manuel Peixoto do Nascimento
Links: Internal Page

Associated Publications
  • 2Boock Chapters
  • J. Nascimento, J. Bioucas-Dias, Hyperspectral Imagery Framework for Unmixing and Dimensionality Estimation, Chapter in, Pattern Recognition - Applications and Methods, Pedro Latorre Carmona, J. Salvador Sánchez, Ana L.N. Fred, Elsevier Science, 2013
  • J. Nascimento, J. Bioucas-Dias, Signal Subspace Identification in Hyperspecral Imagery, Chapter in, Signal and Image Processing for Remote Sensing, C.H. Chen, CRC Press, 2012