In the last few years, the cultural heritage field has posed its attention over image processing techniques, in particular for the diagnostic of art works. Many different images of the same painting are taken, in different parts of the light spectrum; these different images have to be fused together, to get an augmented image showing much more details and information which are only visible in some sub-parts of the spectrum singularly. In this paper, we pose the attention over the compulsory step accomplished just before the "fusion" of such images, i.e. the registration step; in particular, we present an automatic registration technique, based on the computation of mutual information. By means of the registration, it is possible to exactly align the different images, which is the preliminary step to obtain a useful and correct augmented image. These techniques can be also applied to other areas, and in particular to remote sensing images
[1]
Jan Flusser,et al.
An application of image processing in the medieval mosaic conservation
,
2003,
Pattern Analysis and Applications.
[2]
Milan Sonka,et al.
"Handbook of Medical Imaging, Volume 2. Medical Image Processing and Analysis "
,
2000
.
[3]
Guy Marchal,et al.
Automated Multimodality Medical Images Registration using Information Theory
,
1995
.
[4]
Paul Suetens,et al.
Medical image registration using mutual information
,
2003,
Proc. IEEE.
[5]
Guy Marchal,et al.
Multimodality image registration by maximization of mutual information
,
1997,
IEEE Transactions on Medical Imaging.