This paper presents the mathematical framework for the attitude and position estimation of an object from its image using three point correspondence information. A closed-form solution for the pose determination problem is derived by representing the relative attitude parameters using quaternions. The quaternion elements can be used to model the outputs of a vision sensor in attitude control simulations for the final approach phase of a rendezvous and docking mission. The algorithm has been validated with real images of a scaled model of the target features in a vision system test facility. Simulation results showing the comparison between the actual and estimated pose parameters are included.