Pointing the line-of-sight of an acquisition, tracking and pointing (ATP) system at a target requires designation of a reference line of sight (LOS) based principally on the target image. For many system, this will also include registration of a specific fiducial on the target for precision pointing. It is difficult to select a track reference because of algorithm sensitivity to modeling of the image. The selection of a track reference ins further complicated by image variations associated with changes in the viewing geometry and target characteristics. This paper compares several image-processing algorithms for the precision pointing of a near-space ATP platform that is viewing missile targets. The platform has state-of-the-art alignment, stabilization and tracking functions. The algorithms are tested in a full imaging and control system simulation that models an illuminating laser beam, target reflectance, optical effects, the sensor, a high order control system and pointing dynamics. The target models are based on flight dynamics, orientation, measured drawings and surface reflectivity. The simulation results are compared by calculating bias, drift and jitter characteristics of the error incurred when attempting to point the optical line-of- sight at the target. Several algorithms have been identified that provide a pointing reference capable of sustaining sub- microradian error. This paper describes the calculation of the fiducials for the algorithms and compares their relative merits.