Numerical and experimental works are conducted to develop a visualization technique for the phase distribution in a two-phase system by electrical impedance tomography technique. For the two-phase flow system, the impedance of each phase can be known but instead the phase boundary depending on the distribution of dispersed phase is of interest. Here, the image reconstruction problem is derived as a boundary estimation one and a novel algorithm is developed for the estimation of phase boundary. Based on the synthetic and real experimental data, the images of the simulated bubble distributions are reconstructed successfully.