In this paper, the shape reconstruction of a perfectly conducting cylinder buried in a half-space by measured transverse magnetic scattered field and the modified particle swarm optimization (MPSO). Assume that a conducting cylinder of unknown shape is buried in one half-space and scatters the field incident from another half-space where the scattered filed is measured. Based on the boundary condition and the measured scattered field, a set of nonlinear integral equations is derived and the imaging problem is reformulated into an optimization problem. The inverse problem is resolved by an optimization approach, and the global searching scheme PSO is then employed to search the parameter space. Two algorithms: the PSO and the MPSO have been examined. Both techniques have been tested in the case of simulated measurements contaminated by additive Gaussian noise. Numerical results demonstrate that even when the initial guess is far away from the exact one, the two algorithms can both obtain good reconstruction and the MPSO outperforms the PSO in convergence speed.