Quantifying Internal Properties of Geologic Materials by Image Analysis
Previously Published Material: Some were reported at previous AGU meetings. This is intended as an invited review of image analysis techniques in support of the session.
Abstract ID#: 33436
3D information can be obtained using stereological methods on thin section or polished sections, but can only be used for simple convex shapes. For more complex shapes (with concavities, voids, etc.) it is necessary to use fully 3D techniques, such as those provided by Computed X-ray Tomography (CXT). Extracting quantitative information requires measuring the number of voxels within an object of interest. In order to determine the position of the boundary between internal objects (e.g. void vs. glass) “segmentation” of the images is a first step. Oftentimes, the resolution of the x-ray generated voxels is insufficient to resolve thin films between objects, and many small objects are imaged as a single very large, interconnected object. The volumes of the bubbles in pumices before coalescence can thus be reconstructed by sequentially “peeling” away the outer layer, then reconstructing vesicles with individual identities. Once this is accomplished, sizes, shapes, orientations, and proximity to neighbors can all be quantified.
Some geological materials are so small that imaging requires SEM techniques (e.g. ash, clay, etc.). Digital elevation models can be constructed for particles using Stereo SEM images. This is enabled by “dirty” surfaces that enable multi-directional reflection from the surfaces of the particles. For volcanic ash, the sizes of the bubbles that burst during fragmentation to form the ash can be determined by the radii of curvature of concavities on ash surfaces. SSEM can still only image the top of an ash particle, so if a reconstruction of the entire particle is desired, the lower half must be numerically reconstructed from the morphology of the upper half.
In sum, there are various techniques for obtaining quantitative information from a suite of imaging technologies, and specific approaches must typically be developed to answer specific research questions regarding geological materials.