Volumetric Texture Analysis
Texture analysis in 2D has been well studied, but many
3D applications in Medical Imaging,
Stratigraphy or Crystallography, would benefit from
3D analysis instead of the traditional, slice-by-slice approach.
During my PhD at Warwick University I developed a Multiresolution
Volumetric
Texture Segmentation (M-VTS) algorithm which allowed the analysis of
texture through 3D frequency filtering. The method ology extracted
textural measurements from the Fourier domain of the data via subband
filtering using an Orientation Pyramid [1].
A novel Bhattacharyya space, based on the
Bhattacharyya distance was proposed for selecting the most discriminant
measurements and producing a compact feature space. Each dimension of
the feature space was used to form the lowest level of a Quad Tree. At
the highest level of the tree, new positional features were added to
improve the contiguity of the classification. The classified space was
then projected to lower levels of the tree where a boundary refinement
procedure is performed with a 3D equivalent of butterfly filters.
The performance of M-VTS was tested in 2D by classifying
a set of standard texture images. M-VTS yields lower misclassification
rates than reported elsewhere.
The algorithm was tested in 3D with
artificial isotropic data and three Magnetic
Resonance Imaging sets of human knees with encouraging results.
The regions segmented from the knees correspond to anatomical
structures that could be used as a starting point for other
measurements. By way of example, we demonstrate successful cartilage
extraction.
- an in-depth analysis of extraction of textural features from volumetric data will soon apper here:
- the novel Bhattacharyya space and its application in 2D texture
Reyes-Aldasoro, C.C., and A. Bhalerao,
The Bhattacharyya space for feature selection and its application to texture segmentation,
Pattern Recognition, (2006)
Vol. 39, Issue 5, May 2006, pp. 812-826.
- and the volumetric segmentation in 3D MRI images:
Reyes-Aldasoro, C.C., and A. Bhalerao,
Volumetric Texture Segmentation by Discriminant Feature Selection and Multiresolution Classification,
IEEE Trans. on Medical Imaging (2007)
Vol. 25, No. 1, pp. 1-14.
- all the datasets and algorithms are available on an as-is basis in my old website at Warwick:
