Estimation of fiber orientation and length distribution in cashmere fibrous assemblies

In a general fibrous assembly, single fiber orientation as well as fiber length distributions are important characteristics because they directly influence the properties of textiles. An X-ray micro-tomography experiment with a high resolution of 3 μm was for the first time conducted on a randomly oriented inner Mongolia cashmere fibrous assembly to get a series of two-dimensional projections from different angles and the corresponding cone–beam algorithm proposed by Feldkamp and volume rendering technique were used to realize the three-dimensional (3D) reconstruction. An automated segmentation algorithm described by Rigort and Weber was used to trace and detect single fibers from tomographic 3D data. Local normalized cross-correlation of the tomograms was computed with a cylindrical template that mimics a short microtubule segment to get two new objects, named the correlation and orientation fields. Tracing results of fiber length and orientation distributions were given in this paper statistically.