A PSO/GA (particle swarm optimization/genetic algorithm) based method for optimal placement of a stereo-vision measurement system was discussed. With the goal of finding the optimal variables to minimize the uncertainties associated with the reconstruction of 3D points, an objective function for optimal placement of a stereo-vision measurement system was constructed. The procedure for solving the optimal parameters of a stereo vision measurement system was described, and the proposed method was verified using MATLAB. The simulation results show that the curve of best of generation fitness value obtained by PSO/GA hybrid algorithm raises more quickly than those obtained by the traditional particle swarm optimization and genetic algorithm, and the optimal solutions minimize the uncertainties associated with the reconstruction of 3D points.