Techniques for image enhancement and segmentation of tomographic images of porous materials

This article presents a three-stage approach, combining novel and traditional algorithms, for the segmentation of images of porous and composite materials obtained from X-ray tomography. The first stage is an anisotropic diffusion filter which removes noise while preserving significant features. The second stage applies the unsharp mask sharpening filter which enhances edges and partially reverses the smoothing that is often a consequence of tomographic reconstruction. The final stage uses a combination of watershed and active contour methods for segmentation of the grey-scale data. For the data sets we have analysed, this approach gives the highest quality results. In addition, it has been implemented on cluster-type parallel computers and applied to cubic images comprising up to 20003 voxels.