Reconstruction and characterization of the porous composite electrode via experimental and numerical approaches is one of the most important ingredients of mesoscopic modeling. It is also the basis and prerequisite for bottom-to-up design and optimization of electrode microstructure. In the present work, a simulated annealing approach is employed to reconstruct the LiCoO2 cathode of a commercial Li-ion battery. Important statistical characteristic parameters of the real LiCoO2 cathode, such as porosity or component volume fraction, the real size distribution curve of LiCoO2 particles, which are taken from experimental data or extracted from the source materials used to fabricate the cathode, are used to regulate the reconstruction process. The reconstructed electrode evidently distinguishes the three individual phases: LiCoO2 as active material, pores or electrolyte, and additives. An extensive characterization is subsequently performed, which calculates some important structural and transport properties, including the geometrical connectivity of an individual phase, the specific surface area, etc. Particularly, a self-developed D3Q15 LB (lattice Boltzmann) model is utilized to calculate the effective thermal (or electric) conductivity and the effective species diffusivity in electrolyte (or solid) phase, and the tortuosity of an individual phase. The LB model predictions indicate that the effective transport coefficients are closely related to the micro-morphology in electrodes and the tortuosity values assessed by LBM are more reliable than those predicted by random walk simulation or the Bruggeman equation.