Ceramic foams are promising, highly porous materials, with a wide range of specific surface area and low fluid flow resistance, which are well-suited for filtering applications. They are comprised mainly of macrovoids that are interconnected with struts. A branch-shaped reconstruction algorithm is introduced in the present work to reconstruct various ceramic foams from electron microscopy images using the Laguerre tessellation method. Subsequently, the reconstructed samples are used for the numerical calculation of pore structure and transport properties, including specific surface area, tortuosity, effective diffusivity, and flow permeability. Following comparison with experimental data, this reconstruction method is shown to be more reliable than typical analytical expressions that are suggested in the literature for the aforementioned structural and transport properties. Extracting the equivalent pore radius of the reconstructed domains offers improved accuracy of the analytical expressions for the permeability estimation.