This dataset contains time-resolved X-ray micro-CT volumes processed by threshold-based segmentation and ROI extraction, to enable object tracking within large 3D volumes under size constraints. The experiment consists in a yarn of 6 strings, made with hemp fibres. It is fixed in a tube of the tomography holder which allows to add fluid at the base. The tomography was repeated 20 times at the same height and every minutes on the sample after immersing its bottom in water. The goal of this public dataset is to provide code and data that explain how to track the trajectories of fibres. The first step demonstrates how to extract sub-volumes from reconstructed tomographic files, using thresholding and region-of-interest extraction, to obtain the fibre shape using Python code. Once binarized, Paraview scripts are used to gather fibres data at all time steps and identify some ‘recognizable’ fibre in the domain. Connectivity filter allows to build object from ‘adjacent’ cells. The largest objects are kept as ‘recognizable’ fibres, whose movements are represented in figures. The entire process is provided, from raw data extraction to figure generation, in a step-by-step procedure. These time-resolved micro-CT volumes and analysis scripts were designed to monitor the motion of fibres during capillary rise, enabling reproducible 3D tracking of internal structures.

Paraview, 5.11.0

Python, 3.11