A ground-based photo-electronic surveillance system based on passive optical systems was investigated for improving its recognition ability and recognition efficiency for faint targets in mid-high orbits. Meanwhile, an optimized processing method for original images was presented to enhance the Signal to Noise Ratios(SNRs)of the faint targets and improve the system recognition efficiency. The optical reflection characteristics of orbital targets were analyzed and a optimized design scheme for the ground-based photo-electronic surveillance system was given by comparing the SNRs in different modes. In combined with target motion, observation conditions and other factors, an observation model of Geosynchronous Orbit (GEO) targets was designed. Finally, in view of faint target recognition, a new target recognition method for faint targets was presented based on optimization principle. According to the measured data, this method was validated and compared with that of the traditional difference frame method. The results show that the target can be identified when the target SNR is greater than 3.09. This research provides a high reference value for the design and application of photo-electronic surveillance equipment for faint targets in the mid-high orbit.