Raman images are restored using interior point least squares (IPLS). Bounding hyperplane-based hyperspectral unmixing reduces the dimensionality of the data cube, substantially decreasing computation time. IPLS is then independently applied to all resulting map images. The proposed method is first validated on simulated data and then being applied to real Raman spectroscopic data. The results allow for better physical characterisation of the sample and the spatial resolution is enhanced from 400nm to better than 50nm.

This presentation is part of Contributed Presentation “CP8 - Contributed session 8”

Authors:

Dominik J. Winterauer (Renishaw plc,; Institut des Matériaux Jean Rouxel)

Said Moussaoui (Laboratoire des Sciences du Numérique de Nantes UMR CNRS 6004, Ecole Centrale de Nantes )

Tim Batten (Spectroscopy Products Division, Renishaw plc)

Bernard Humbert (Institute des Materiaux Jean Rouxel Nantes, University of Nantes)

Daniel Funes-Hernando (Institut des Matériaux Jean Rouxel)

Keywords:

hyperspectral image restoration, image deblurring, image enhancement, inverse problems, nonlinear optimization