We
all use color in our daily lives to make important decisions. Our
natural color vision is based on three sensors, generally well suited
to making distinctions in many natural scenes. Hyperspectral imaging
allows us to extend color imaging to multiple wavelength bands, thus
making finer distinctions. It finds applications in the military,
in airborne monitoring of the health of coral reefs, in manufacturing,
and in medical imaging.
The
research team at Northeastern University and the University of Puerto
Rico at Mayaguez, has built a capability for tabletop hyperspectral
imaging, with an 8-nanometer spectral resolution across the visible
spectrum. Application areas have included medicine and remote sensing.
Research
Activities:
This
study will involve collecting hyperspectral images of small natural
scenes, skin, or man-made objects. The research team will use MATLAB
and other software packages to collect and analyze data. Techniques
such as principal component analysis will be employed. Teachers
will gain an understanding of light propagation in complex media,
digital optical imaging, development of models and algorithms, and
their implementation.