Exploring the relationship between root and soil has significance in environmental protection. But it has long been a challenge to see through the soil structure and investigate the root’s growth without destroying its local environment. To analyze the interaction between soil and root during its growth, we grew plants in a granular medium consisting of spherical particles and used X-ray computed tomography (CT) for visualization. The experiments for identifying the optimal parameters for CT measurement were carried out, providing a systematic method for future experiments. The algorithms for detecting the particles’ coordinates and extracting the root surface from the tomogram were modified based on the previous work. The particle detection algorithm was proven to have excellent accuracy, while the root extracting algorithm requires further improvement. Voro++ and Pomelo were used to generate the (Set-)Voronoi tessellation of our system, and the packing fractions of particles and roots were calculated. The local packing fraction distribution of particles was consistent with our theory, but data around the root surface needed further analysis. Overall, this work’s workflow of experiments, image processing, and structural analysis can serve as an instruction for future systematic measurements of root growth in granular media using X-ray CT.