|
Interdisciplinary Workshop on
|
 EVAN Network (EC) |
 Museum of Toulouse |
 ODENT Project, CNRS |
| Presentation | Organization | Introduction | Session 1 | Session 2 | Session 3 | Participants | Minutes | Media |
New CT-Scan systems allow to obtain 3D images with an increasing resolution.
In particular, Micro-CT imaging devices give 3D images with a resolution of 10-20 microns wich can reach up to 400 nanometers [Ritman, 2004]. Many companies propose such equipment for clinical (e.g. [Scanco], [Skyscan], [GEHealthcare]) or industrial (e.g. [Viscom], [Xradia]) applications. This modality has been used since several years to get accurate details of fossil structures (see for example [Thompson & Illerhaus, 1999]) and several paleo-anthropology laboratories have bought micro-CT scanners (e.g. [MPIEA] or [UTCT]) or collaborate with specific institutions (e.g. [MEDES]).
But if more and more micro-CT images are available, a lot of problems has arose. First, there is no standard acquisition protocol as in classical medical CT-imaging (see for example [Badawi-Fayad et al., 2005]). Moreover, it seems that the quality of the images can depend on the device itself [Olejniczak et al., 2007]. Research work should be done in order to define the image quality and to compare it among different protocols and devices. It is particularly important in the case of fossil which are embedded in rock as some artifacts may appear.
Another difficulty is the size of micro-CT data. For example, the file of an image composed of 1000 slices of 1000x1000 pixels is more than 1.8 gigabyte large! It becomes very complex to store and process such data on a standard PC [Semal et al., 2005]. New algorithms for segmentation, filtering, registration and visualization have to be developed in order to decompose the image into smaller sub-images which can be processed either sequantially or in parallel by several processors. We can also use multiresolution methods