X-ray tomography is a powerful technique that allows us to see very tiny details inside a rock. However, the image acquisition is usually just a starting point for the image analysis. In order to get quantifiable information, one has to develop specific image processing algorithms. In the porous medium research, one of the most important processing step is the development of the task oriented image segmentation algorithm.
While trying our segmentation algorithm on a 3D image of a sedimentary rock, we found some curious piece of a former life! The “worm” you can see in the video is an orthoceras — an ancient mollusk that is often found in sediments.
This carbonate rock has been cored in the Miocene carbonate platform of Llucmajor, in Majorca. The rock has suffered a re-equilibration from aragonite to calcite (dissolution of aragonite and crystallization of calcite). This reaction led to the formation of porosity (grey parts of the picture). In this case the spatial distribution of the pores has been controlled by the pre-existing structure of the rock. This process allowed the preservation of the shape of the fossil, even after re-equilibration and recrystallization into calcite. That is why we can see the orthoceras, although its skeleton has undergone chemical alteration.
Figure 1. This video is a series of 2D slices of a 3D volume. No orthoceras is actually swimming here
If you are interested in fossils, you have probably seen orthoceras before: polished orthoceras are often used as souvenirs and are widely found on market. The special thing of the orthoceras from our video is that it is just 2.5mm long, when the polished souvenir orthoceras are typically 3-20 cm long.
Figure 2. Polished orthoceras (source: lvmnh.org)
By the way, the segmentation algorithm we were testing performed very well.
Working automatically, it segmented the image into eight phases of different porosity. Although the quality of the segmentation could be improved with some manual adjustment of a few parameters, it is still an excellent result for an automatic algorithm.
Figure 3. Result of segmentation. Black color correspond to solid, white to voids, and grey colors correspond to phases of intermediate porosity