Abstract: In dispersed fuel pellets, core particles are randomly dispersed and distributed in non-metallic materials. The method of characterizing the position information of core particles is usually used to evaluate and evaluate the quality of product processes, which is of great significance. This paper adopted the three-dimensional CT image method and used VG software to fit the outer contour of the core block to obtain the three-dimensional coordinates of the core block. The core particles were treated as inclusions, and on the basis of obtaining the center coordinates of each core particle, the distance between adjacent core particles was automatically calculated. Then, microfocal CT was used to detect and process dispersed fuel pellets, and statistical heat map was used to visually characterize the distribution uniformity of core particles in the fuel pellets. Finally, the automatic measurement of the spacing between adjacent core particles was tested using actual fuel pellets, proving the feasibility of this method for measuring the spacing between core particles. This lays a technical foundation for further characterization of core particle distribution.